THE NEW REALISM

THE MACMILLAN COMPANY .

NEW YORK • BOSTON • CHICAGO
DALLAS • SAN FRANCISCO

MACMILLAN & CO., LIMITED

LONDON • BOMBAY • CALCUTTA
MELBOURNE

THE MACMILLAN CO. OF CANADA, LTD.

TORONTO

THE NEW REALISM

COOPERATIVE STUDIES IN PHILOSOPHY

BY

EDWIN B. HOLT

WALTER T. MARVIN

WILLIAM PEPPERRELL MONTAGUE

RALPH BARTON PERRY

WALTER B. PITKIN

AND

EDWARD GLEASON SPAULDING

THE MACMILLAN COMPANY
1912

All rights reserved

COPYRIGHT, 1912,
BY THE MACMILLAN COMPANY.

Set up and electrotyped. Published August, 1912.

Norinooti

J. 8. Cashing Co. — Berwick & Smith Co.
Norwood, Mass., U.S.A.

Ul

-1VERSITY OF CALIFORNIA
SANTA BARBAI^ ^. u.LEGE LIBRARY

PREFACE

ON July 21, 1910, we published a brief article entitled ' The Pro-
gram and First Platform of Six Realists,' * in which we indicated the
direction philosophical inquiry ought to take. We there asserted
that advance would be facilitated by cooperative investigations ; and
the drafting of the platform was a first attempt to confirm this
belief. The present volume continues, on a larger scale, the work
there inaugurated ; and we hope it will be followed by other col-
lections of studies.

The introductory essay voices our common opinions. The other
essays do so only in part. It has seemed best to publish them with-
out laboring for complete unanimity, inasmuch as their agreements
quite overshadow their differences. They have been written after
prolonged conferences. A few important debatable topics are briefly
discussed by dissenting members in the Appendix.

DECEMBER 31, 1911.

1 J. of Phil., Psychol., etc., 7, 393. This is reprinted in the Appendix.

TABLE OF CONTENTS

INTRODUCTION

SECTION PAGE

I. THE HISTORICAL SIGNIFICANCE OF THE NEW REALISM ... 2

1. Naive realism, 2 — 2. Dualism, 4 — 3. Subjectivism of the
Berkeleian type, 5 — 4. Subjectivism of the Kantian type, 8 —
5. The new realism, 9.

II. THE REALISTIC POLEMIC ......... 11

1. The fallacy of argument from the ego-centric predicament,
11 — 2. The fallacy of pseudo-simplicity, 12 — 3. The fallacy of
exclusive particularity, 14 — 4. The fallacy of definition by initial
predication, 15 — 5. The speculative dogma, 16 — 6. The error of
verbal suggestion, 18 — 7. The fallacy of illicit importance, 19.

in. THE REALISTIC PROGRAM OF REFORM ...... 21

1. The scrupulous use of words, 21 — 2. Definition, 22 —

3. Analysis, 24 — 4. Regard for logical form, 25. — 5. Division
of the question, 26 — 6. Explicit agreement, 28 — 7. The separa-
tion of philosophical research from the study of the history of
philosophy, 30.

IV. REALISM AS A CONSTRUCTIVE PHILOSOPHY 31

1. Implications of the rejection of subjectivism, 32 — 2. Impli-
cations of the rejection of anti-intellectualism, 32 — 3. Monism
and pluralism, 33 — 4. Knowledge and its object; the independ-
ence of the object, 33 — 6. Identity of content and thing known,
34 — 6. Platonic realism, 35 — 7. Summary, 36.

V. REALISM AND THE SPECIAL SCIENCES ...... 36

1. The general attitude of realism to the special sciences, 36
— 2. Realism and psychology, 37 — 3. Realism and biology, 39 —

4. The relation of realism to logic and the mathematical sci-
ences, 40 — 6. Realism as a basis for cooperation, 41.

THE EMANCIPATION OF METAPHYSICS FEOM

EPISTEMOLOGY
BY WALTER T. MARVIN

I. THE ISSUE BETWEEN DOGMATISM AND CRITICISM .... 45

1. Epistemology regarded as a science logically prior to all other
sciences, 45 — 2. Epistemology regarded as a science of the limits

viii TABLE OF CONTENTS

SECTION PACK

and possibility of knowledge, 46 — 3. Epistemology regarded as a
theory of reality, 47 — 4. Summary, 49 — 5. The propositions
held by the dogmatist in opposition to criticism, 49 — 6. The con-
clusion which this essay will endeavor to establish, 50.

II. THE THEORY OF KNOWLEDGE NOT LOGICALLY FUNDAMENTAL . 51

1. Two errors suspected to be present in the argument of the
criticist, 51 — 2. Logic is not a science of the laws of thought, 52
— 3. The subject matter of logic, 52 — 4. This subject matter is
non-mental, 52 — 5. Logic is not the art of correct thinking, 53 —
6. The way in which we use logic in our thinking, 54 — 7. Sum-
mary, 65 — 8. Ambiguity of the word ' knowledge ' : the knowing
process and the thing known, 56 — 9. The subsistence of proposi-
tions, 57 — 10. Conclusion, 60.

III. THE LOGICAL POSITION, RELATIVELY TO THE OTHER SCIENCES, OF

THE THEORY OF KNOWLEDGE AND IN PARTICULAR OF THE
PROBLEM OF THE POSSIBILITY OF KNOWLEDGE ... 60
1. The science of the possibility of knowledge presupposes logic,
60 — 2. The doctrine that the science of the possibility of knowl-
edge is fundamental to all sciences other than logic, 61 — 3. Science
itself and belief in the propositions of science do not presuppose
the possibility of knowledge, 62 — 4. An ultimate premise is, as
such, beyond investigation, 63 — 5. A restatement of the problem,
67 — 6. The theory of knowledge is logically subsequent to many
of the special sciences, 67 — 7. So also is the science of the limits
and possibility of knowledge, 70 — 8. Conclusion, 73.

IV. EPISTEMOLOGY DOES NOT GIVE, BUT PRESUPPOSES A THEORY OF

REALITY 74

1. Two types of transcendentalism, 74 — 2. Objections of the
dogmatist to the first type of transcendentalism, 77 — 3. A brief
examination of Kantian transcendentalism to illustrate these objec-
tions, 78 — 4. Objections to the second type of transcendentalism :
(a) reality a self-consistent system, 80 — 5. (ft) reality an organic
unity, 82 — 6. Conclusion, 82.

V. AN APPEAL TO THE PRAGMATIC TEST, TO THE VERDICT OF HISTORY 83

1. To what has the change in our modern conception of the
world been due? 83 — 2. The influence of mathematics, 85 —
3. The influence of physics, 86 — 4. The influence of the doctrine
of evolution and of historical research, 86 — 5. Influences at work
in the doctrine of primary and secondary qualities, 87 — 6. Influ-
ences at work in the doctrine of causation, 89 — 7. Influences at
work in the doctrine of substance, 90 — 8. Influences at work in
spiritualism and monistic absolutism, 91 — 9. Conclusion, 91.

TABLE OF CONTENTS ix

SECTION PAGE

VI. METAPHYSICS SHOULD BE EMANCIPATED FROM EPISTEMOLOGY . . 92

1. The logical and methodological indebtedness of metaphysics
to other bodies of knowledge, 92 — 2. Its indebtedness to episte-
mology, 94 — 3. General conclusion, 94.

A REALISTIC THEORY OF INDEPENDENCE
BY EALPH BARTON PERRY

I THE IMPORTANCE OF THE NOTION OF INDEPENDENCE ... 99

1. Its meaning in the older realism, 99 — 2. How used by Reid
and Locke, 100 — 3. The resulting confusion of realism with sub-
stantialism, 103 — 4. The present necessity of clarifying the con-
cept, 104.

II. MEANINGS OF THE TERM DEPENDENCE ...... 106

1. Relation, 106 — 2. Whole-part, 107—3. Part-whole, 107 —
4. Thing-attribute, 109 — 5. Attribute-thing, 109 — 6. Causation,
109 — 7. Reciprocity, 111 — 8. Implying, 112 — 9. Being implied,
112.

III. THE MEANING OF INDEPENDENCE IN NEO-REALISM . . . .113

1. Independence not non-relation, 113 — 2. Independence not
priority, 115 — 3. Independence denned, 117.

IV. A REALISTIC THEORY OF INDEPENDENCE FORMULATED IN GENERAL

TERMS 118

1. All simples mutually independent, 118 — 2. And independent
of complexes of which they are members, 119 — 3. Complexes
mutually independent, 119 — 4. Complexes dependent on their sim-
ple constituents, 119 — 5. One case of a complex depending upon
another, 119 — 6. Another case, 120 — 7. A third case, 121 — 8.
A fourth case, 121 — 9. When one complex is independent of an-
other, 122 — 10. How one entity may acquire dependence, 124.

V. A REALISTIC THEORY OF INDEPENDENCE APPLIED TO THE CASE OF

KNOWLEDGE 126

1. An experienced entity related to a complex, 126 — 2. Simple
entities not dependent on consciousness, 126 — 3. Complexes inde-
pendent of knowledge as respects their simple constituents, 129 —
4. Logical and mathematical propositions independent of con-
sciousness, 129 — 5. Physical complexes independent of conscious-
ness, 130 — 6. Logical, mathematical, and physical complexes as
objects of consciousness, 132.

VL CASES OF SUBJECTIVITY, OR DEPENDENCE ON A PRIMARY CONSCIOUS-
NESS 136

1. Parts of consciousness dependent on the whole, 137 — 2. Lim-
ited reciprocal dependence of parts of consciousness, 137 —

TABLE OF CONTENTS

3. Some elements dependent on selective action, 138 — 4. And
some on combining action, 139 — 5. Value dependent on con-
sciousness, 140 — 6. Works of art, 141 — 7. History, society,
life, and reflective thought, 142.

VII. THE INDEPENDENCE OF SUBJECTIVITY ON A SECONDARY CONSCIOUS-
NESS ..........'. 144

1. The subject of consciousness independent of being known,
144 — 2. One consciousness independent of another, 145 —
3. Mental content independent of introspection, 147 — 4. Value
independent of value-judgments, 148 — 5. Perception independ-
ent of reflection, 149.

VIII. CONCLUSION 151

A DEFENSE OF ANALYSIS

BY EDWARD GLEASON SPAULDING

I. INTRODUCTORY 155

1. The types of analysis and of wholes, 155 — 2. What is
analysis ? 167.

II. COLLECTIONS AND ENUMEBATIVE ANALYSIS ..... 162

HI. THE SECOND TYPE OF WHOLE (SPACE, TIME, ETC.) AND ITS

ANALYSIS 169

1. Arithmetical analysis, 173 — 2. Numbers, 173 — 3. Rational
positive integers, 173 — 4. Relations, 175 — 5. Rational frac-
tions, 177 — 6. Irrational and real numbers, 178 — 7. The analy-
sis of space, 181 — 8. The analysis of time, 190 — 9. Motion and
its analysis, 193 — 10. Velocity and acceleration, 204 — 11. Ac-
celeration, 209 — 12. Dynamics and duration, 212 — 13. Other
classes of individuals (atoms, etc.), 225.
IV. PERCEPTUAL AND CONCEPTUAL ANALYSIS 230

V. THE ANALYSIS OF ORGANIC WHOLES 237

1. Chemical compounds, etc., 237 — 2. Organisms and their
analysis, 243.

A REALISTIC THEORY OF TRUTH AND ERROR
BY WILLIAM PEPPERRELL MONTAGUE

. I. THE MEANING OF TRUTH AND ERROR ...... 252

1. Definition of true and false, 252 — 2. The meaning of real
and unreal, 252 — 3. Objections to the definitions; the verbal

TABLE OF CONTENTS xi

SECTION PACK

fallacy of psychophysical metonymy, 256 — 4. First consequences
of the verbal fallacy of psycbophysical metonymy, 257 — 5. Sec-
ond consequences of the verbal fallacy of psychophysical me-
tonymy, 260 — 6. Summary, 262.

II. CAUSALITY AND CONSCIOUSNESS IN A WORLD OF PURK FACT . . 263

1 Space, time, and quality as the ultimates of factual analysis,
263 — 2. The antinomy of causality — substantist thesis and posi-
tivist antithesis, 264 — 3. The antinomy of consciousness — pan-
hylist antithesis and panpsychist thesis, 268 — 4. A supplementary
antinomy of consciousness ; are perceived objects inside or outside
the brain? 276 — 5. Consciousness and causality; hylopsychism
as the reciprocal solution of the two antinomies, 278 — 6. The
three directions of a potentiality, 281 — 7. A certain difficulty in
terminology, 282 — 8. The three levels of potentiality, 283 —
9. Summary, 285.

III. THE GENESIS OF TRUTH AND ERROR ...... 286

1. The epistemological triangle, 286 — 2. The two kinds of truth
and error, 289 — 3. Attention and belief, 292 — 4. The material
fallacy of psychophysical metonymy, 294 — 5. Degrees of truth
and error and the fallacy of internal relations, 297 — 6. Sum-
mary, 300.

THE PLACE OF ILLUSORY EXPERIENCE IN A
REALISTIC WORLD

BY EDWIN B. HOLT

I. ILLUSIONS OF PERCEPTION AND THOUGHT ...... 303

1. Errors of space, 303 — 2. Errors of time, 307 — 3. Errors in
secondary qualities, 308. — 4. Illusions of thought, 365.

II. ERROR 357

1. Images assert nothing, 357 — 2. What the realist asserts, 358
— 3. Contradiction and being, 360 — 4. Contradiction and real-
ism, 367.

SOME REALISTIC IMPLICATIONS OF BIOLOGY

BY WALTER B. PITKIN
I. THE BIOLOGICAL ATTACK ON REALISM 378

II. FORMAL ANALYSIS OF THE BIOLOGICAL SITUATION .... 380

1. Simple description of some stimuli and reactions, 382 —
A. Descriptive difference between stimuli and reactions, 382 —

xii TABLE OF CONTENTS

BECTION PA.GK

B. Some types of reaction, 386 — a. Simple adjustment, 389

— b. Simple selection, 393 — c. Conduction, 396; the flatfish,
397 — d. Transformation, 405 — e. Resistances, 409 — i. Resist-
ance by destruction of stimulus, 409 — ii. By interpolation, 411

— iii. By suspense of collateral functions, 412.

HI. RESULTS OF THIS FORMAL ANALYSIS ...... 414

1. Perception non-constitutive, 416 — 2. No internal relations
in organisms, 422 — 3. The total situation not reacted to, 424 —
4. Pure relations as stimuli, 425 — The fallacy of indiscernibles,
428.

IV. SOME DEFECTS IN MODERN THEORIES OF CONSCIOUSNESS . . 434

1. The morphological fallacy, 434 — 2. The pragmatic point of
view, 437 — 3. The new realistic analyses, 438 — 4. The biological
point of view, 442.

V. THE BIOLOGICAL STATUS OF CONSCIOUSNESS 443

1. The structure of the environment, 443 — 2. The environment
as a space complex, 445 — 3. The factors of the biological situa-
tion, 453 — 4. Projective indiscernibles, 458 — 5. Hallucinatory
objects, 461 — 6. Conclusion, 466.

APPENDIX

Program and first platform of six realists, 471 — Montague on Holt, 480 —
Holt on Montague, 482 — Pitkin on Montague and Holt, 483.

INDEX . . 487

THE NEW REALISM

INTRODUCTION »

THE new realism may be said to be at the present moment some-
thing between a tendency and a school. So long as it was recog-
nized only by its enemies it was no more than a tendency. But
war has developed a class-consciousness, .and the time is near at
hand, if indeed it is not already here, when one realist may recog-
nize another. This dawning spirit of fellowship, accompanied
by a desire for a better understanding and a more effective co-
operation, has prompted the present undertaking.

It is perhaps inevitable that the new realism should for a time
remain polemical in tone. A new philosophical movement in-
variably arises as a protest against tradition, and bases its hope of
constructive achievement on the correction of established habits
of thought. Neo-realism is still in a phase in which this critical
motive dominates, and is the chief source of its vigor and unanim-
ity. Before, however, a philosophy can come of age, and play a
major part in human thought, it must be a complete philosophy,
or must at least show promise of completeness. If it is to assume
the role, it must undertake to play the whole part. The authors
of the present book thus entertain the hope that they may have
succeeded not only in amplifying, clarifying, and fortifying the
realistic critique, but also in exhibiting that critique as a basis for
the solution of special philosophical problems, and for the pro-
cedure of the special sciences.

1 The following introduction expresses opinions common to the several authors
of this book; but it has proved convenient to make use of parts of the following
articles which have already appeared in print. Montague. The New Realism
and the Old. /. of Phil., Pyychol, etc., 1912, 9, 39. Perry. Realism as a Polemic
and Program of Reform. J. of Phil., Pnychol., etc., 1910, 7, 337, 365.
B X

INTRODUCTION

THE HISTORICAL SIGNIFICANCE OF THE NEW REALISM

THE new realism is not an accident, nor a tour de force, nor an
isolated and curious speculative eruption. Whatever may be
thought of its correctness or power to endure, it must at least be
accorded a place in the main current of modern thought. It is a
fundamental and typical doctrine — definable in terms of the
broad play of intellectual forces, and peculiarly characteristic of
their present conjunction.

The historical significance of the new realism appears most
clearly in its relations with 'naive realism/ 'dualism' and 'sub-
jectivism.' The new realism is primarily a doctrine concerning
the relation between the knowing process and the thing known;
and as such it is the latest phase of a movement of thought which
has already passed through the three phases just indicated. Neo-
realism, in other words, seeks to deal with the same problem that
has given rise to 'naive realism,' 'dualism' and 'subjectivism';
and to profit by the errors as well as the discoveries for which these
doctrines have been responsible.

1. The theory of naive realism is the most primitive of these
theories. It conceives of objects as directly presented to con-
sciousness and being precisely what they appear to be. Nothing
intervenes between the knower and the world external to him.
Objects are not represented in consciousness by ideas; they are
themselves directly presented. This theory makes no distinction
between seeming and being ; things are just what they seem. Con-
sciousness is thought of as analogous to a light which shines out
through the sense organs, illuminating the world outside the
knower. There is in this naive view a complete disregard of the
personal equation and of the elaborate mechanism underlying
sense perception. In a world in which there was no such thing as
error, this theory of the knowledge relation would remain unchal-
lenged; but with the discovery of error and illusion comes per-

HISTORICAL SIGNIFICANCE OF NEW REALISM 3

plexity. Dreams are perhaps the earliest phenomena of error to
arouse the primitive mind from its dogmatic realism. How can a
man lie asleep in his bed and at the same time travel to distant
places and converse with those who are dead ? How can the events
of the dream be reconciled with the events of waking experience ?
The first method of dealing with this type of error is to divide the
real world into two realms, equally objective and equally external,
but the one visible, tangible, and regular, the other more or less
invisible, mysterious, and capricious. The soul after death, and
sometimes during sleep, can enter the second of these realms. The
objectified dreamland of the child and the ghostland of the sav-
age are the outcome of the first effort of natural realism to cope
with the problem of error. It is easy to see, however, that this
doubling up of the world of existing objects will only explain a
very limited number of dream experiences, while to the errors of
waking experience it is obviously inapplicable. Whenever, for
example, the dream is concerned with the same events as those al-
ready experienced in waking life, there can be no question of ap-
pealing to a shadow world. Unreal events that are in conflict
with the experience of one's fellows, and even with one's own more
inclusive experience, must be banished completely from the ex-
ternal world. Where, then, shall they be located ? What is more
natural than to locate them inside the person who experiences
them ? For it is only upon him that the unreal object produces any
effect. The objects of our dreams and our fancies, and of illusions
generally, are held to exist only 'in the mind.' They are like
feelings and desires in being directly experienced only by a single
mind. Thus the soul, already held to be the mysterious principle
of life, and endowed with peculiar properties, transcending ordi-
nary physical things, is further enriched by being made the habitat
of the multitudinous hosts of non-existent objects. Still further
reflection on the phenomena of error leads to the discovery of the
element of relativity in all knowledge, and finally to the realiza-
tion that no external happening can be perceived until after it has
ceased to exist. The events we perceive as present are always

4 INTRODUCTION

past, for in order to perceive anything it must send energy of some
kind to our sense organs, and by the time the energy reaches us
the phase of existence which gave rise to it has passed away. To
this universal and necessary temporal aberration of perceived
objects is added an almost equally universal spatial aberration.
For all objects that move relatively to the observer are perceived
not where they are when perceived, but, at best, where they were
when the stimulus issued from them. And in addition to these
spatial and temporal aberrations of perception we know that what
we perceive will depend not only upon the nature of the object but
on the nature of the medium through which its energies have passed
on their way to our organism ; and also upon the condition of our
sense organs and brain. Finally, we have every reason to believe
that whenever the brain is stimulated in the same way in which it
is normally stimulated by an object we shall experience that ob-
ject even though it is in no sense existentially present. These
many undeniable facts prove that error is no trivial and excep-
tional phenomenon, but the normal, necessary, and universal
taint from which every perceptual experience must suffer.

2. It is such considerations as these that have led to the aban-
donment of naive realism in favor of dualism, the second of the
aforementioned theories. According to this second theory, which is
exemplified in the philosophies of Descartes and Locke, the mind
never perceives anything external to itself. It can perceive only
its own ideas or states. But as it seems impossible to account for
the order in which these ideas occur by appealing to the mind in
which they occur, it is held to be permissible and even necessary
to infer a world of external objects resembling to a greater or less
extent the effects, or ideas, which they produce in us. What we
perceive is now held to be only a picture of what really exists.
Consciousness is no longer thought of as analogous to a light which
directly illumines the extra-organic world, but rather as a painter's
canvas or a photographic plate on which objects in themselves
imperceptible are represented. The great advantage of the second
or picture theory is that it fully accounts for error and illusion;

HISTORICAL SIGNIFICANCE OF NEW REALISM 5

the disadvantage of it is that it appears to account for nothing else.
The only external world is one that we can never experience, the
only world that we can have any experience of is the internal world
of ideas. When we attempt to justify the situation by appealing
to inference as the guarantee of this unexperienceable externality,
we are met by the difficulty that the world we infer can only be
made of the matter of experience, that is, can only be made up of
mental pictures in new combinations. An inferred object is al-
ways a perceptible object, one that could be in some sense experi-
enced ; and, as we have seen, the only things that according to this
view can be experienced are our mental states. Moreover, the
world in which all our interests are centered is the world of experi-
enced objects. Even if, per impossibile, we could justify the belief
in a world beyond that which we could experience, it would be but
a barren achievement, for such a world would contain none of the
things that we see and feel. Such a so-called real world would
be more alien to us and more thoroughly queer than were the ghost-
land or dreamland which, as we remember, the primitive realist
sought to use as a home for certain of the unrealities of life.

3. It seems very natural at such a juncture to try the experi-
ment of leaving out this world of extra-mental objects, and con-
tenting ourselves with a world in which there exist only minds and
their states. This is the third theory, the theory of subjectivism.
According to it, there can be no object without a subject, no exist-
ence without a consciousness of it. To be, is to be perceived.
The world of objects capable of existing independently of a knower
(the belief in which united the natural realist and the dualistic
realist) is now rejected. This third theory agrees with the first
theory in being epistemologically monistic, that is, in holding to the
presentative rather than to the representative theory of percep-
tion ; for, according to the first theory, whatever is perceived must
exist, and according to the present theory, whatever exists must be
perceived. Naive realism subsumed the perceived as a species
under the genus existent. Subjectivism subsumes the existent as
a species under the genus perceived. But while the third theory

6 INTRODUCTION

has these affiliations with the first theory, it agrees with the second
theory in regarding all perceived objects as mental states — ideas
inhering in the mind that knows them and as inseparable from
that mind as any accident is from the substance that owns it.

Subjectivism has many forms, or rather, many degrees. It oc-
curs in its first and most conservative form in the philosophy of
Berkeley. Descartes and Locke, and other upholders of the
dualistic epistemology, had already gone beyond the requirements
of the picture theory in respect to the secondary qualities of ob-
jects. Not content with the doctrine that these qualities as they
existed in objects could only be inferred, they had denied them even
the inferential status which they accorded to primary qualities.
The secondary qualities that we perceive are not even copies of
what exists externally. They are the cloudy effects produced in
the mind by combinations of primary qualities, and they resemble
unreal objects in that they are merely subjective. The chief
ground for this element of subjectivism in the systems of dualis-
tic realism immediately preceding Berkeley, was the belief that
relativity to the percipient implied subjectivity. As the secondary
qualities showed this relativity, they were condemned as subjec-
tive. Now it was the easiest thing in the world for Berkeley to
show that an equal or even greater relativity pertained to the
primary qualities. The perceived form, size, and solidity of an
object depend quite as much upon the relation of the percipient
to the object as do its color and temperature. If it be axiomatic
that whatever is relative to the perceiver exists only as an idea,
why, then, the primary qualities which were all that remained of
the physical world could be reduced to mere ideas. But just here
Berkeley brought his reasoning to an abrupt stop. He refused
to recognize that (1) the relations between ideas or the order in
which they are given to us, and (2) the other minds that are known,
are quite as relative to the knower as are the primary and secondary
qualities of the physical world. You can know other minds only in
so far as you have experience of them, and to infer their independent
existence involves just as much and just as little of the process of

HISTORICAL SIGNIFICANCE OF NEW REALISM 7

objectifying and hypostatizing your own ideas as to infer the in-
dependent existence of physical objects. Berkeley avoided this
obvious result of his own logic by using the word 'notion' to
describe the knowledge of those things that did not depend for
their existence on the fact that they were known. If you had an
idea of a thing — say of your neighbor's body — then that thing
existed only as a mental state. But if you had a notion of a thing
— say of your neighbor's mind — then that thing was quite ca-
pable of existing independently of your knowing it. Considering
the vigorous eloquence with which Berkeley inveighed against
the tendency of philosophers to substitute words for thoughts,
it is pathetic that he should himself have furnished such a striking
example of that very fallacy. In later times Clifford and Pearson
have not hesitated to avail themselves of a quite similar linguistic
device for escaping the solipsistic conclusion of a consistent sub-
jectivism. The distinction between the physical objects which
as ' constructs ' exist only in the consciousness of the knower, and
other minds which as 'ejects' can be known without being in any
way dependent on the knower, is essentially the same both in its
meaning and in its futility as the Berkeleian distinction of idea
and notion. For the issue between realism and subjectivism does
not arise from a psychocentric predicament — a difficulty of con-
ceiving of objects apart from any consciousness — but rather from
the much more radical 'ego-centric predicament,'1 the difficulty
of conceiving known things to exist independently of my knowing
them. And the poignancy of the predicament is quite independ-
ent of the nature of the object itself, whether that be a physical
thing such as my neighbor's body, or a psychical thing such as my
neighbor's mind.

Some part of this difficulty Hume saw and endeavored to meet
in his proof that the spiritual substances of Berkeley were them-
selves mere ideas; but Hume's position is itself subject to two
criticisms : First, it succeeds no better than Berkeley's in avoid-
ing a complete relativism or solipsism — for it is as difficult to ex-

»Cf. below, 11-12.

8 INTRODUCTION

plain how one 'bundle of perceptions' can have any knowledge
of the other equally real 'bundle of perceptions' as to explain how
one 'spirit' can have knowledge of other 'spirits.' Second, the
Humean doctrine suffers from an additional difficulty peculiar to
itself, in that by destroying the conception of the mind as a 'sub-
stance,' it made meaningless the quite correlative conception of
perceived objects as mental 'states.' If there is no substance
there cannot be any states or accidents, and there ceases to be
any sense in regarding the things that are known as dependent upon
or inseparable from a knower.

4. Passing on to that form of subjectivism developed by Kant,
we may note three points: (1) A step back toward dualism, in
that he dallies with, even if he does not actually embrace, the
dualistic notion of a ding-an-sich, a reality outside and beyond the
realm of experienced objects which serves as their cause or ground.
(2) A step in advance of the subjectivism of Berkeley and Hume,
in that Kant reduces to the subjective status not merely the facts
of nature but also her laws, so far, at least, as they are based upon
the forms of space and time and upon the categories. (3) There
appears in the Kantian system a wholly new feature which is des-
tined to figure prominently in later systems. This is the dualistic
conception of the knower, as himself a twofold being, tran-
scendental and empirical. It is the transcendental or noumenal
self that gives laws to nature, and that owns the experienced ob-
jects as its states. The empirical or phenomenal self, on the other
hand, is simply one object among others, and enjoys no special
primacy in its relation to the world of which it is a part.

The post-Kantian philosophies deal with the three points just
mentioned in the following ways: (1) The retrograde feature of
Kant's doctrine — the belief in the ding-an-sich — is abandoned.
(2) The step in advance — the legislative power conferred by Kant
upon the self as knower — is accepted and enlarged to the point
of viewing consciousness as the source not only of the a priori forms
of relation, but of all relations whatsoever. (3) The doctrine of
the dual self is extended to the point of identifying in one absolute

HISTORICAL SIGNIFICANCE OF NEW REALISM 9

self the plurality of transcendental selves held to by Kant, with
the result that our various empirical selves and the objects of their
experience are all regarded as the manifestations or fragments of
a single, perfect, all-inclusive, and eternal self. But it is not
hard to see that this new dualism of the finite and the absolute
selves involves the same difficulties as those which we found in the
Cartesian dualism of conscious state and physical object. For
either the experience of the fragment embraces the experiences of
the absolute or it does not. If the former, then the absolute be-
comes knowable, to be sure, but only at the cost of losing its ab-
soluteness and being reduced to a mere 'state ' of the alleged frag-
ment. The existence of the absolute will then depend upon the
fact that it is known by its own fragments, and each fragmentary
self will have to assume that its own experience constitutes the
entire universe — which is solipsism. If the other horn of the
dilemma be chosen and the independent reality of the absolute be
insisted upon, then it is at the cost of making the absolute unknow-
able, of reducing it to the status of the unexperienceable external
world of the dualistic realist. The dilemma itself is the inevitable
consequence of making knowledge an internal relation and hence
constitutive of its objects. Indeed, a large part of the philosophi-
cal discussion of recent years has been concerned with the endeavor
of the absolutists to defend their doctrine from the attacks of
empiricists of the Berkeleian and Humean tradition in such a way
as to avoid equally the Scylla of epistemological dualism and the
Charybdis of, solipsism. But, as we have seen, the more empirical
subjectivists of the older and strictly British school are open to
the same criticism as that which they urge upon the absolutists;
for it is as difficult for the Berkeleian to justify his belief hi the
existence of other spirits, or the phenomenalistic follower of Hume
his belief in bundles or streams of experience other than his own,
as for the absolutist to justify those features [of the absolute
experience which lie beyond the experience of the finite frag-
ments.

5. And now enter upon this troubled scene the new realists,

10 INTRODUCTION

offering to absolutists and phenomenalists impartially their new
theory of the relation of knower to known.

From the standpoint of this new theory all subjectivists suffer
from a common complaint. The ontological differences that
separate such writers as Fichte and Berkeley, Mr. Bradley and
Professor Karl Pearson, are, for a realist, overshadowed by the
epistemological error that unites them. The escape from subjec-
tivism and the formulation of an alternative that shall be both reme-
dial and positively fruitful, constitutes the central preeminent issue
for any realistic protagonist. It is prior to all other philosophical
issues, such as monism and pluralism, eternalism and temporalism,
materialism and spiritualism, or even pragmatism and intellec-
tualism. This does not mean that the new realism shall not lead
to a solution of these problems, but only that as a basis for their
clear discussion it is first of all essential to get rid of subjec-
tivism.

The new realists' relational theory is in essentials very old.
To understand its meaning it is necessary to go back beyond Kant,
beyond Berkeley, beyond even Locke and Descartes — far back
to that primordial common sense which believes in a world that
exists independently of the knowing of it, but believes also that
that same independent world can be directly presented in con-
sciousness and not merely represented or copied by 'ideas.' In
short, the new realism is, broadly speaking, a return to that naive
or natural realism which was the first of our three typical theories
of the knowledge relation ; and as such, it should be sharply dis-
tinguished from the dualistic or inferential realism of the Car-
tesians. But the cause of the abandonment of naive realism in
favor of the dualistic or picture theory was the apparently hope-
less disagreement of the world as presented in immediate experi-
ence with the true or corrected system of objects in whose reality
we believe. So the first and most urgent problem for the new
realists is to amend the realism of common sense in such wise as
to make it compatible with the facts of relativity.

For this reason especial attention has been given in the present

THE REALISTIC POLEMIC 11

volume 1 to a discussion of those special phenomena, such as illu-
sion and error, which are supposed to discredit natural realism,
and set going a train of thought that cannot be stopped short of
subjectivism. It is necessary to inquire closely into the mechan-
ism of perception, and into the logic of contradiction and falsity.
And it is necessary to obtain a definition of the central thesis of
realism, the thesis of independence, that shall not be so loose as
to violate the facts, nor so vague and formal as to disregard them.2

II

THE REALISTIC POLEMIC

INASMUCH as subjectivism, renewed and fortified under the name
of 'idealism/ is the dominant philosophy of the day, it affords
the chief resistance which an innovating philosophy such as realism
has to overcome. The realistic polemic is therefore primarily
a polemic against subjectivism; but the errors of which realism
finds subjectivistic philosophies to be guilty, are not necessarily
confined to such philosophies. They may be generalized; and
in so far as they are generalized their discovery is of greater mo-
ment. The following are some of the traditional errors which
neo-realism has thus far succeeded in generalizing.

1. The fallacy of argument from the ego-centric predicament.
— The 'ego-centric predicament' consists in the impossibility of
finding anything that is not known.3 This is a predicament rather
than a discovery, because it refers to a difficulty of procedure,
rather than to a character of things. It is impossible to eliminate
the knower without interrupting observation ; hence the peculiar
difficulty of discovering what characters, if any, things possess
when not known. When this situation is formulated as a proposi-
tion concerning things, the result is either the redundant inference
that all known things are known, or the false inference that all

1 Cf. below, Nos. IV, V, VI. * Cf. below, No. II.

1 In this connection, ' known ' means ' given as an object of thought.'

12 INTRODUCTION

things are known. The former is, on account of its redundancy,
not a proposition at all; and its use results only in confusing it
with the second proposition, which involves a petitio principii.
The falsity of the inference, in the case of the latter proposition,
lies in its being a use of the method of agreement unsupported by
the method of difference. It is impossible to argue from the fact
that everything one finds is known, to the conclusion that knowing
is a universal condition of being, because it is impossible to find
non-things which are not known. The use of the method of agree-
ment without negative cases is a fallacy. It should be added that
at best the method of agreement is a preliminary aid to exact
thought, and can throw no light whatsoever on what can be
meant by saying that knowing is a condition of being. Yet this
method, misapplied, is the main proof, perhaps the only proof, that
has been offered of the cardinal principle of idealistic philosophies
— the definition of being in terms of consciousness. It is difficult,
on account of their very lack of logical form, to obtain pure cases of
philosophical fallacies. Then, too, this particular fallacy has so far
become a commonplace as to be regarded as a self-evident truth.
The step in which it is employed is omitted or obscured in many
idealistic treatises. In others it is spread so thin, is so pervasive
and insidious, that while it lends whatever support is offered for
the cardinal idealistic principle, it is nowhere explicitly formulated.
But the following will serve as a typical illustration. "Things
exist," says Renouvier, "and all things have a common character,
that of being represented, of appearing ; for if there were no repre-
sentation of things, how should I speak of them?" 1 It is clear
that no more is proved by this argument than that things must be
'represented' if one is to 'speak of them.' That all things have
the common character of being 'spoken of,' which is the funda-
mental thesis restated in a new form, is left without any proof
whatsoever.

2. The fallacy of pseudo-simplicity. — There is a disposition
in philosophy as well as in common sense to assume the simplicity

1 Renouvier. Mind, 1877, 2, 378.

THE REALISTIC POLEMIC 13

of that which is only familiar or stereotyped. This error has
conspired with the error just examined to lend a certain plausi-
bility to subjectivism. For one would scarcely assert with so
much gravity that the world was his idea, or that the 'I think*
must accompany every judgment, unless he supposed that the
first personal pronoun referred to something that did not require
further elucidation. Self-consciousness could never have figured
in idealistic philosophies as the immediate and primary certainty
if it were understood to be a complex and problematic conception.
Yet such it must be admitted to be, once its practical sim-
plicity, based on habits of thought and speech, is discounted.
Similarly the common dogma, to the effect that consciousness can
be known only irrespectively, is based on the assumption that it
is known introspectively, and that thus approached it is a simple
datum. Traditional spiritistic conceptions of will, activity, im-
mediacy, and life, rest on the same fundamental misapprehension
as does the materialistic acceptance of body as an irreducible en-
tity. Thus what is really at stake here is nothing less than the
method of analysis itself. In exact procedure it is not permitted
to assert the simplicity of any concept until after analysis. That
the concepts enumerated above are not analytically simple, is
proved by the fact that when they are treated as simple, it is
necessary to give them a complex existence also in order to account
for what is known about them. It is customary to say that this
is a 'manifestation' or 'transformation' of the simple and more
fundamental reality; but this is to reverse the order which is
proper to thought as the deliberate and systematic attempt to
know. It is equivalent to asserting that the more pains we take
to know, the less real is the object of our knowledge ; a proposition
which is never asserted without being contradicted, since it ex-
presses the final critical analysis of the thinker who asserts it. The
following is a characteristic example of the error of 'pseudo-sim-
plicity,' as applied to the conception of activity.

"Every man," says Professor Ward, "knows the difference be-
tween feeling and doing, between idle reverie and intense thought,

14 INTRODUCTION

between impotent and aimless drifting and unswerving tenacity
of purpose, being the slave of every passion or the master of him-
self. ... It must surely ever remain futile, nay, even foolish,
to attempt to explain either receptivity or activity; for what is
there in experience more fundamental? And being thus funda-
mental, the prime staple of all experience, it is absurd to seek to
prove them real, since in the first and foremost sense of reality the
real and they are one." 1 Nevertheless, activity and passivity
are capable of being analyzed in a variety of ways, logical, physical,
and psychological ; 2 and their nature can be regarded as a simple
datum only in so far as such analysis is deliberately avoided. They
are simples only in so far as they are not yet analyzed,

3. The fallacy of exclusive particularity. — It is ordinarily as-
sumed that a particular term of any system belongs to such system
exclusively. That this is a false assumption is proved empirically.
The point 6 of the class of points that constitutes the straight line
abc may belong also to the class of points that constitutes the inter-
secting straight line xby. The man John Doe who belongs to the
class Republican Party may belong also to the intersecting class
captains of industry. Unless this multiple classification of terms
were possible, discourse would break down utterly. All the terms
of discourse are general in the sense that they belong to several
contexts. It is this fact that accounts for the origin and the
usefulness of language. Without this generality of terms the
world would possess no structure, not even motion or similarity;
for there could be no motion if the same could not be in different
places at different times, and there could be no similarity if the
same could not appear in different qualitative groupings. It is
little wonder, then, that the virtual rejection of this principle by
philosophy has led to a fundamental and perpetual difficulty. To
this error may perhaps be traced the untenability of Platonic
universalism, recognized apparently by Plato himself, and the

1 Ward, J. Naturalism and Agnosticism, 2, 52, 53.

» Cf. e.g. James, W. The Experience of Activity, in Essays in Radical Em-
piricism, VI.

THE REALISTIC POLEMIC 15

untenability of modern particularism, attested by the desperate
efforts which almost every modern philosopher has made to save
himself from it.

The most familiar variety of particularism is found in naturalism.
This may be traced to the naive bias for the space-time order, or
that historical series of bodily changes which constitutes the course
of nature. Naturalism asserts that this is the only system, and
that its terms, the several bodily events, belong to it exclusively.
That this theory is untenable is evident at once, since in order that
bodily events shall possess the structure and connections necessary
to them, being must contain other terms, such as places, times,
numbers, etc., that are not bodily events. But historically,
naturalism has been discredited mainly by its failure to provide
for the system of ideas, a system without which the bodily system
itself could not be known ; and it is the exclusive particularity of
the terms of this latter that has figured most prominently in
philosophical discussions.

In dualism of the Cartesian type the terms of nature and the
terms of knowledge are regarded as exclusive, but in order that
knowledge shall mean anything at all, it is assumed that there is
some sort of representative relation between them. Spinoza and
Leibniz endeavored to bring them together through a third and
neutral term. Among the English philosophers the impossibility
of showing how the mind can know nature if each mind is a closed
circle, possessing its content wholly within itself, leads finally to
the abolition of nature as an independent system. Thus the pen-
dulum swings from naturalism to subjectivism; and in the whole
course of this dialectic the mistaken principle of exclusive par-
ticularity is assumed.

4. The fallacy of definition by initial predication. — This form
of error is a natural sequel to the last. A subject of discourse is
viewed initially under one of its aspects, or is taken initially as a
term in some specific complex or relational manifold. Then, owing
to the error of exclusive particularity, it is assumed that this sub-
ject of discourse can have no other aspect, or belong to no other

16 INTRODUCTION

relational manifold. Thus the initial characterization becomes
definitive and final.

Subjectivism, again, affords the most notable instances of the
error. Any subject of discourse may be construed as such; that
is, as a thing talked about or 'taken account of,' as an object of
experience or knowledge. The vogue of the psychological, in-
trospective, or reflective method in modern thought has given rise
to the custom of construing things first according to their place in
the context of consciousness. Similarly, the habit of self-con-
sciousness among philosophers has emphasized the relation of
things to self; and the prominence of epistemology in modern
philosophy has tended to an initial characterization of things ac-
cording to their places in the process of knowledge, just as the
prominence of religious issues led early Christian ascetics to
name things first after their part in the drama of the soul's
salvation.

Thus, idealism, quite unconscious of having prejudged the main
question from the outset, "seeks to interpret the universe after
the analogy of conscious life, and regards experience as for us the
great reality."1 Or, as another writer expressed it, "we must
start . . . from the whole of experience as such." 2 But all such
initial characterizations must be regarded as accidental. Allow-
ance must be duly made for alternative and complementary char-
acterizations; and the question of the priority of the characteri-
zation to which any subject of discourse submits must be discussed
quite independently of the order which is determined by habit or
bias. In short, the very general disposition at the present time
to begin with a psychological or epistemological version of things
must not be allowed in the least to prejudice the question as to
whether that version is definitive or important.

5. The speculative dogma. — By the 'speculative dogma' is
meant the assumption for philosophical purposes that there is
an all-sufficient, all-general principle, a single fundamental propo-

1 Lindsay, J. Studies in European Philosophy, 207.

* Baillie, J. B. Idealistic Construction of Experience, 105.

THE REALISTIC POLEMIC 17

sition that adequately determines or explains everything. This
assumption has commonly taken one or the other of two forms.
By many it has been assumed that such a principle constitutes
the proper content or subject matter of philosophy. Thus Plato
said: "And when I speak of the other division of the intelligible
you will understand me to speak of that other sort of knowledge
which reason herself attains by the power of dialectic, using the
hypotheses not as first principles, but only as hypotheses — that is
to say, as steps and points of departure into a region which is above
hypotheses, in order that she may soar beyond them to the first
principle of the whole ; and clinging to this and then to that which
depends on this, by successive steps she descends again without the
aid of any sensible object, beginning and ending in ideas." l And
Caird makes the same assumption when he says that "Philosophy
professes to seek and to find the principle of unity which underlies
all the manifold particular truths of the separate sciences." 2 But
such an assumption is dogmatic, because it ignores the prior ques-
tion as to whether there is such a principle or not. So far as the
general task of philosophy is concerned, this must be treated as an
open question. Philosophy does aim, it is true, to generalize as
widely and comprehend as adequately as possible; but a loosely
aggregated world, abounding in unmitigated variety, is a philo-
sophical hypothesis. The discovery of a highly coherent system
under which all the wealth of experience could be subsumed would
be the most magnificent of philosophical achievements; but if
there is no such system, philosophy must be satisfied with some-
thing less — with whatever, in fact, there happens to be. By
others, in the second place, it has been assumed that the idea of
such a principle or system is the property of every thoughtful per-
son, the existence of an object corresponding to it being alone
doubtful. This assumption gave rise to the ontological proof of
God, which carried conviction only so long as man did not question
the definiteness and meaning of the idea; for the assumption

1 Plato, (Jowett, trans.) Republic, 511, B.
* Caird, E. The Social Philosophy and Religion of Comte, xiii.
0

18 INTRODUCTION

obscured a problem, the problem, namely, as to whether there is
any idea corresponding to the words ens realissimum. The pos-
sibility of defining, on general logical grounds, a maximum of being
or truth, is, to say the least, highly questionable; and it is cer-
tain that this problem must properly precede any inferences from
such a maximal idea.

The speculative dogma has been the most prolific cause of the
verbal abuses which abound in philosophy, and which are to be
considered separately. It is through this dogma that various
words have been invested with a certain hyperbole and equivo-
cation, in consequence of the attempt to stretch their meaning
to fit the speculative demand. A further evil arising from the
speculative dogma is the unjust and confusing disparagement of
positive knowledge through invidious comparison with this Un-
known God to which the philosopher has erected his altar.

6. The error of verbal suggestion. — Words which do not possess
a clear and unambiguous meaning, but which nevertheless have a
rhetorical effect owing to their associations, lend themselves to a
specious discourse, having no cognitive value in itself, and stand-
ing in the way of the attainment of genuine knowledge. This is
Bacon's famous idol of the forum. In philosophy this reliance on
the suggestive, rather than the proper denotative or connotative
function of words, is due not only to man's general and ineradicable
tendency to verbalism, but also to the wide vogue of doctrines
that are fundamentally inarticulate. We have already examined
two errors which lead philosophers to accept such doctrines. The
error of pseudo-simplicity involves a reference to topics that
cannot be analytically expressed; they cannot be identified and
assigned an unequivocal name. The speculative dogma has, as
we have seen, led to the use of words which shall somehow convey
a sense of finality, or of limitless and exhaustive application, where
no specific object or exact concept possessing such characters is
offered for inspection. This is what Berkeley calls the "method
of growing in expression, and dwindling in notion." Ordinarily
the words so used have a precise meaning also, and there results a

THE REALISTIC POLEMIC 19

double evil. On the one hand, the exact meaning of such
terms as 'force,' 'matter,' 'consciousness,' 'will/ etc., is blurred
and vitiated; and on the other hand, their speculative meaning
borrows a content to which it is not entitled. The desire of philos-
ophers to satisfy the religious demand for an object of worship or
faith, doubtless one of the fundamental motives of the speculative
dogma, leads to yet another variety of verbal suggestion, in which
a technical philosophical conception is given a name that possesses
eloquence and power of edification. Thus philosophers commonly
prefer the term 'eternal' to the term 'non-temporal,' and 'in-
finite' to 'series with no last term,' or 'class, a part of which
can be put in one-to-one correspondence with the whole.' Such
terms as ' significance,' ' supreme,' ' highest,' ' unity,' have a
similar value. Or the same end may be achieved by decorating
almost any word with a capital letter, as is exemplified by the
emotional difference between truth and Truth, or absolute and
Absolute.

Finally, there is a verbal abuse which is worse, even, than
equivocation; for it is possible to invent utterly fictitious con-
cepts simply by combining words. In such cases, the constituent
concepts, if the words happen to signify any, are not united. They
may be positively repugnant, or simply irrelevant. At any rate,
they have not been tested for consistency, and whether they do or
do not constitute a true system or complex concept remains wholly
problematic. Such, for example, is the case with Eucken's "total
activity, which by its own movement develops into an independent
reality and at the same time comprehends the opposition of subject
and object, subjectivity and objectivity. " * Such procedure is the
principal source of the fallacy of obscurum per obscurius and affords
an almost unlimited opportunity for error.

7. The fallacy of illicit importance. — This is one of the most
insidious errors which has ever been foisted upon mankind, and it
is the idealist who has popularized it. It consists in inferring
that, because a proposition is self-evident or unchallengeable, there-

1 Eucken. (Pogson, trans.) Life of the Spirit, 329.

20 INTRODUCTION

fore it is important. There is a healthy animal instinct behind
the fallacy. Men have early learned that the certain affords,
on the whole, a safer basis for conduct than the uncertain. The
merchant who is sure of his market grows rich faster than his igno-
rant competitor. The statesman who is sure of his constituents
acts with directness and decision. So it is throughout all practi-
cal life. Now, the practical man never reflects upon his own men-
tal processes, and thus he fails to note that the certainty he feels
toward things is not an attribute of them, but only a certain pre-
cision in his attitude toward them. But the fact that the relations
are unequivocal and clear is no proof that they happen to be of much
significance. A may surely be C, and yet its being C may be the
most trivial circumstance. A man, for instance, may be abso-
lutely sure he likes cucumbers; but this does not prove that
cucumbers are the true foundation of dietetics, nor that his
liking of them reveals either his own nature or the nature of
cucumbers.

Undeterred by such obvious cases, however, the idealist is
wont to reason that all philosophy and all science must be built
upon the one fact that nobody can make any unchallengeable as-
sertion about anything except his having an immediate experience.

One might ask the idealist whether he is any more certain of
being aware than he is of the presented object ; whether, for ex-
ample, in addition to saying: "I am certain that I am experi-
encing" — he cannot say with equal assurance : "There certainly
is a tree of some sort over yonder." But to take up this debate is
to pass beyond the fallacy which he has committed. And no so-
lution of the question alters the fact that he has erred logically hi
holding that, because A is undeniably B, therefore B is an impor-
tant characteristic of A. There is no sure connection between the
axiomatic and the significant. To think there is, is vicious in-
tellectualism. The fallacy is curable only by the use of strict
logic, but by this very easily. If one person is certain that a dis-
tant object is a tree, while his companion is equally certain that
the same object is an automobile, is it not obvious that certainty

THE REALISTIC PROGRAM OF REFORM 21

is a negligible factor in the problem of deciding what the object
really is?

Ill

THE REALISTIC PROGRAM OF REFORM

PHILOSOPHY has repeatedly thrown off its bad habits, and
aroused itself to critical vigilance. Furthermore, there is good
ground for asserting that there has never before been so great
an opportunity of reform. Logic and mathematics, the tradi-
tional models of procedure, are themselves being submitted to a
searching revision that has already thrown a new light on the gen-
eral principles of exact thinking; and there is promise of more
light to come, for science has for all time become reflectively
conscious of its own method. The era of quarrelsome misun-
derstanding between criticism and positive knowledge is giving
way to an era of united and complementary endeavor. It must
not be forgotten that philosophy is peculiarly dependent on logic.
Natural science in its empirical and experimental phases can safely
be guided by instinct, because it operates in the field of objects
defined by common sense. But the very objects of philosophy
are the fruit of analysis. Its task is the correction of the cate-
gories of common sense, and all hope of a profitable and valid re-
sult must be based on an expert critical judgment. The present
situation, then, affords philosophy an opportunity of adopting a
more rigorous procedure and assuming a more systematic form.
It is with reference to this opportunity that it is worth while
here to repeat the advice which is our common inheritance from
the great philosophical reformers. None of these canons is origi-
nal, but all are pertinent and timely.

1. The scrupulous use of words. — This is a moral rather than
a logical canon. There is need in philosophy of a greater fastidi-
ousness and nicety in the use of words. A regard for words is,
in philosophy, the surest proof of a sensitive scientific conscience ;
for words are the instruments of philosophical procedure, and

22 INTRODUCTION

deserve the same care as the lancet of the surgeon or the balance
of the chemist. A complacent and superior disregard of words is
as fatuous as it is offensive. It is a healthier intellectual symptom
to feel as Maclan felt in Chesterton's 'Ball and the Cross.'
"Why shouldn't we quarrel about a word? What is the good of
words if they aren't important enough to quarrel over ? Why do
we choose one word more than another if there isn't any difference
between them ? If you called a woman a chimpanzee instead of
an angel, wouldn't there be a quarrel about a word? If you're
not going to argue about words, what are you going to argue about ?
Are you going to convey your meaning to me by moving your ears ?
The church and the heresies always used to fight about words,
because they are the only things worth fighting about." l

2. Definition. — "The light of human minds," says Hobbes,
"is perspicuous words, but by exact definitions first snuffed and
purged from all ambiguities." Words are properly signs. They
are serviceable in proportion as they are self-effacing. A skillful
word will introduce the hearer or reader to his object, and then
retire ; only the awkward word will call attention to itself. It fol-
lows, then, that the only means of escaping quarrels about words
is to use words with discrimination, with careful reference to their
objective purport, or usefulness as means of access to ideas. Fur-
thermore, a word is essentially a social instrument, whether used
for record or communication, and requires that its relation to an
object or idea shall be agreed on and conventionalized. This is
the only means of bringing several minds together in a common
topic of discourse. "Syllables," says John Toland, "though never
so well put together, if they have not ideas fix'd to them, are but
words spoken in the air, and cannot be the ground of a reasonable
service." 2

Philosophy is peculiarly dependent upon a clear definition of
the reference of words because, as we have already seen, its objects
are not those of common sense. It cannot rely on the ordinary

1 Chesterton. The Ball and the Cross, 96.

1 Toland. Christianity not Mysterious (2d ed.) , 30.

THE REALISTIC PROGRAM OF REFORM 23

denotation of words. This fact affords a perennial and abundant
source of confusion, from which there is no escape save through
the creation of a technical vocabulary. Bacon's observations on
this matter are worthy of being quoted in full. "Now words,"
he says, "being commonly framed and applied according to the
capacity of the vulgar, follow those lines of division which are most
obvious to the vulgar understanding. And whenever an under-
standing of greater acuteness or a more diligent observation would
alter those lines to suit the true divisions of nature, words stand
in the way and resist the change. Whence it comes to pass that
the high and formal discussions of learned men end oftentimes in
disputes about words and names, with which (according to the
use and wisdom of the mathematicians) it would be more prudent
to begin, and so by means of definitions reduce them to order. " l
Definition, then, means, in the first instance, the unequivocal
and conventional reference of words. But there is a further ques-
tion which arises from the use of single words to refer to complex
objects. If such a reference is to be unequivocal, it is necessary
that there should be a verbal complex mediating between the single
word and the complex object. Thus if a circle is defined as 'the
class of points equidistant from a given point,' this means that a
circle is a complex object whose components are specified by the
words in the given phrase. The single word is virtually an abbre-
viation of the phrase. The clarity of words depends in the end
on their possessing a conventional reference to simple objects.
But with the progress of analysis and the demonstration of the
unsuspected or unexplored complexity of things, the single word
which at first denoted the object in its pre-analytical simplicity,
comes to stand for several words which denote the components of
the object in their post-analytical simplicity. Definition, then,
means two things : first, a convention regarding the substitution
of a single word for a group of words ; second, a convention re-
garding the reference of words to objects.2

1 Bacon. Novum Organum (edition of Ellis and Spedding), IV, 61.

2 The definition of things, rather than words, is apparently the same as knowledge
in general.

24 INTRODUCTION

3. Analysis. — The term 'analysis' properly refers not to the
special method of any branch of knowledge, but to the method
of exact knowledge in general, to that method of procedure in
which the problematic is discovered to be a complex of simples.
Such procedure may lead to the discovery of fine identities in the
place of gross differences, or fine differences in the place of gross
identities. Analysis in this sense means only the careful, sys-
tematic, and exhaustive examination of any topic of discourse.
It cannot, then, be proper to assert that such procedure destroys
its object. It does, it is true, require that naivete" and innocence
of mind shall give place to sophistication ; or that ignorance shall
give place to some degree of explicitly formulated knowledge. But
even the discovery that such psychological or moral values are
lost is itself the result of analysis. Nor is there any difficulty in
providing a place for such values within the psychological or moral
systems to which they belong. In the second place, it cannot be
proper to assert that there is anything which necessarily escapes
analysis, such as 'real' change or 'real' activity. The method
of analysis does not require that change and activity shall be any-
thing other than what any investigation shall discover them to be.
Analysis may show either that they are unanalyzable or that they
may be further reduced. If they turn out to be unanalyzable, it
can only be because they exhibit no complexity of structure, no
plurality of necessary factors. If they turn out to be reducible,
then they must be identical with the totality of their components.
If they appear to differ from such a totality, then they must appear
so to differ in some respect, and this respect must at once be added
to complete the totality. It is especially important not to forget
the combining relations. A toy is not identical with the collection
of the fragments into which it has been shattered, but it is identi-
cal with these fragments in that particular arrangement which has
been destroyed. Similarly dynamics does not reduce motion to
the occupancy of positions, but to the occupancy of positions in a
temporal order. There is a perfectly clear difference between
geometry or statics, on the one hand, and dynamics on the other.

THE REALISTIC PROGRAM OF REFORM 23

It is important also not to confuse analysis and synthesis with the
physical operation that often accompanies them. For the pur-
poses of knowledge it is not necessary to put Humpty Dumpty
together again, but only to recognize that Humpty Dumpty is not
himself unless the pieces are together.

The common prejudice against analysis is due in part to this
false supposition that it is an attempt to substitute a collection of
parts for an arrangement of parts. But it is due also to a more or
less habitual confusion between things and words. Those who
have employed the analytical method have been by no means
guiltless in the matter. So soon as any word obtains currency it
begins to pose as a thing in its own right, and discourse is con-
stantly tending to take on the form of a logomachy. It has not
unnaturally been supposed that analysts intended to verbalize
reality, to give to its parts the artificial and stereotyped character
of words, and to its processes the formal arrangement of grammar.
But, as we have already seen, verbalism cannot be avoided by a
deliberate carelessness in the use of words. If words are to be
both useful and subordinate, it is necessary that they should be
kept in working order, like signposts kept up to date, with their
inscriptions legible and their pointing true.

4. Regard for logical form. — Logic is at the present time in a
state of extraordinary activity, and able both to stimulate and to
enrich philosophy. The principal contribution which modern
logic is prepared to make to philosophy concerns the form of exact
knowledge. This problem is by no means wholly solved, and there
is an important work to be done which only philosophers can do.
But the mathematical logicians have already broken and fer-
tilized the ground. The theory of relations, the theory of ' logical
constants' or indefinables, the theory of infinity and continuity,
and the theory of classes and systems, concern everything fun-
damental in philosophy. No philosopher can ignore these and
like theories without playing the part of an amateur. The mathe-
matical logicians may be quite mistaken, or they may have failed
to go to the root of things ; but in that case they must be over-

26 INTRODUCTION

taken in their error and corrected on their own grounds, if the field
of scientific philosophy is not to be abandoned to them altogether.
The present situation is certainly intolerable ; for philosophy
deals with the same topics as modern logic, but treats popularly
and confusedly what modern logic treats with the painstaking
thoroughness and exactness of the expert.

There is another respect in which modern logic should be of
service to philosophy. In the course of a reconstruction of the
foundations of mathematics, certain general canons of good think-
ing have come to light ; and these are directly applicable to philo-
sophical procedure.1 We refer to such canons as 'consistency' and
'simplicity.' These canons are new in the sense that they are
now well enough defined to afford a means of testing any theory.
A theory is consistent when its fundamental propositions actually
generate terms, or when a class can be found which they define ;
and a theory satisfies the criterion of simplicity or parsimony when
none of its fundamental propositions can be deduced from the
rest. It behooves philosophy, then, both to ally itself with logic,
in the investigation of the most ultimate concepts, such as relation,
class, system, order, indefinable, etc., and also to apply to its own
constructive procedure the most refined tests of scientific form.
It is one of the major purposes of the new realism to justify and
to extend the method of logic and of exact science in general. For
this reason one of the essays in this volume 2 is especially devoted
to defending the truthfulness of that method and giving it full
ontological validity.

5. Division of the question. — Although philosophy is especially
charged with correcting the results obtained in each special in-
vestigation by results obtained from other investigations, it is
folly to ignore the necessity, humanly speaking, of dealing with
one problem at a time. Not only is the attempt to raise and
answer all questions together futile, but it prevents either definite-
ness of concepts or cogency of reasoning. Exact knowledge must

1 Cf. Schmidt. Critique of Cognition and its Principles, J. of Phil., Psychol., etc.,
1909, 6, 281. » No. III.

THE REALISTIC PROGRAM OF REFORM 27

be precisely limited in its application. A disposition in philosophy
to employ terms in an unlimited sense, and to make unlimited
assertions, is the principal reason why philosophy at the pres-
ent time possesses no common body of theory. And for the same
reason philosophy is to-day without any common plan of work to be
done. English and American philosophers have been much exer-
cised during the past decade over what is called 'the problem of
truth.' It is assumed that the various parties to this discussion
are referring to the same thing; but it is doubtful if this would
e-rer be suspected, did they not specifically mention one another's
•lames and writings. These quarrels are perhaps due less to dis-
agreement on the merits of any question, than to an irritable de-
termination to be heard. If a sober and patient attempt were
made to reduce the present differences of philosophical opinion
to debatable propositions, the first result would be a division of
the question at issue. It would certainly appear that the present-
day problem of truth is one problem only so long as it is a symbol
of factional dispute ; discuss it, and it at once proves to be many
problems, as independent of one another as any problems can be.
If one undertakes to enumerate these problems, one readily finds
as many as seven : (1) The problem of non-existence : What dis-
position is to be made of negated propositions, of non-temporal
propositions, and of imaginary propositions? (2) The problem
of the one and the many : How may many elements belong to one
system ? (3) The problem of logical form : What are the ultimate
categories? (4) The problem of methodology: How shall one
best proceed in order to know ? (5) The problem of universality :
How can that which is known at a moment transcend that mo-
ment ? (6) The problem of the values of knowledge : What are
the criteria of right believing? (7) The problem of the relation
between belief and its object : In what respect does belief directly
or indirectly modify its object?

If agreement, or even intelligent disagreement, is to be obtained,
philosophical issues must be sharpened. If any steady advance is
to be made, special problems must be examined in order, and one

28 INTRODUCTION

at a time. There is a large group of such special problems that is
by general consent assigned to philosophy. In addition to those
already enumerated, there are such problems as consciousness,
causality, matter, particularity and generality, individuality,
teleology, all of them problems whose solution is of the first im-
portance both for the special sciences and for religious belief.
These problems are examined by the traditional philosophy ; but
they are not sufficiently isolated, nor examined with sufficient
intensive application. They find their place in most philosophical
treatises as applications of a general system, and not as problems
to be examined independently on their merits.

6. Explicit agreement. — The recent discussion of the desir-
ability and expediency of a 'philosophical platform' has developed
a difference of opinion as to whether agreement should be explicit
or implicit.1 Agreement of some sort is conceded to be a desidera-
tum, but there are some who believe that a common tradition or
historical background is all that is necessary. Now is it not evi-
dent that in theoretical or scientific procedure there is no agree-
ment until it is explicitly formulated ? The philosophical classics
afford no basis for agreement, because they are open to interpre-
tation. The difficulty is merely complicated through the necessity
of first agreeing on the meaning of a text. To employ terms and
propositions in their historical sense is to adopt precisely the course
which is adopted by common sense. It means the introduction
into what is supposed to be exact discourse of the indeterminate
human values with which tradition is incrusted. In exact dis-
course the meaning of every term must be reviewed ; no stone can
be allowed to go into the building that has not been inspected and
approved by the builder. Otherwise the individual philosopher
is no more than an instrument in the hands of the welt-geist. He
must be possessed by a fatalistic confidence that the truth will
take care of itself if he only repeats the formulas that he has learned
in the schools or in the market place. But the most precious and

1 Cf. Schmidt, Creighton, and Leighton, /. of Phil., Psychol., etc., 1909, 6, 141,
240, 519, 673.

THE REALISTIC PROGRAM OF REFORM 29

cherished privilege of philosophy is the critical independence of
each generation. Every philosophical reformer from the begin-
ning of European thought has been moved by a distrust of tradi-
tion, and has proclaimed the need of a perpetual watchful-
ness lest the prestige of opinion be mistaken for the weight of
evidence.

If agreement is to be based on tradition, then tradition, with all
its ambiguity, its admixture of irrelevant associations, and its un-
lawful authority, is made the arbiter of philosophical disputes.
That no theoretical difference is ever really judged in this way is
abundantly proved from the present situation in philosophy. We
sympathize, but we do not agree ; we differ, but we do not disagree.
It is of more importance in theoretical procedure that two or three
should agree, than that all should sympathize. "If the trumpet
gives an uncertain sound," says Toland, "who shall prepare him-
self to the battle?" Agreement and disagreement alike require
the explicit formulation of theories in terms freshly defined. It
is not to be supposed that those who insist on the necessity of
explicit agreement have in mind any general unanimity. The
principle would be satisfied if a single philosopher could be found
to agree with himself — provided the agreement were explicit.
For then it would be possible for others to disagree with him, and
to disagree explicitly. We should then have before us a number of
carefully formulated propositions, which could be tested and de-
bated in the light of the evidence, propositions which would be
the common property of philosophers and the material with which
to construct an impersonal system of philosophical knowledge.

The first duty of philosophers, then, is not to agree, but to make
their implicit agreements or disagreements explicit. Moreover it is
not easy to see how this duty can be escaped without entirely
abandoning philosophy's claim to be a theoretical discipline.
If we cannot express our meaning in exact terms, in terms that we
are willing should stand as final, if like the sophists of old we must
make long speeches and employ the arts of rhetoric ; then let us at
least cultivate literature. At present we are bad scientists and

30 INTRODUCTION

worse poets. But philosophy is not necessarily ineffable.1 The dif-
ficulties which some philosophies have in meeting the demands of
exact discourse are gratuitous, and are due to a habit of mixing
theory, on the one hand, with the history of theory, and, on the
other hand, with common belief. It is not necessary that phi-
losophy should abandon its interest in either history or common
belief, but it is necessary that it should isolate those interests, and
not permit them to compromise its direct study of problems.

7. The separation of philosophical research from the study of
the history of philosophy. — A problem can be solved only by the
attentive examination of that which the problem denotes. But a
problem of historical exegesis, and an original philosophical prob-
lem, necessarily denote different things and direct the attention
to different quarters. Thus the problem of Hume's conception of
causality directs attention to a text, whereas the problem of causal-
ity directs attention to types of sequence or dependence exhibited
in nature. It is worth while to formulate this commonplace be-
cause there is a present-day habit of procedure that obscures it.
It is customary to assume that it is the mark of rigorous scholar-
ship in philosophy to confine oneself to commentaries on the
classics. To raise the question of the importance of the history
of philosophy is not necessary. That it has an indispensable place
in human culture and in the discipline of every philosopher is not
to be doubted ; but that it has a higher dignity than a direct and
independent analysis of special problems seems to be nothing more
than a superstition. What dignity the history of philosophy pos-
sesses it derives from the originality of the individual philosophers
whose achievements it records. If philosophy were to consist in
the study of the history of philosophy, it would have no history.
Doubtless the by-product of originality is charlatanry and sopho-
moric conceit ; but mankind is not less well served by this than by
the complacent pedantry which is the by-product of erudition.

But whether the historical form of treatment does or does not

»Cf. Sheffer, H. M. Ineffable Philosophies, /. of Phil, Psychol., etc., 1909, 6,
123.

REALISM AS A CONSTRUCTIVE PHILOSOPHY 31

lend dignity to philosophical discourse, it certainly adds com-
plexity and difficulty. Ferrier, good Hegelian though he was at
heart, confided to his readers the hopelessness of undertaking to
show whether his conclusion agreed with Hegel's or not. "It is
impossible to say to what extent this proposition coincides, or does
not coincide, with his opinions ; for whatever truth there may be
in Hegel, it is certain that his meaning cannot be wrung from him
by any amount of mere reading, any more than the whisky which
is in bread . . . can be extracted by squeezing the loaf into a
tumbler. He requires to be distilled, as all philosophers do more
or less — but Hegel to an extent which is unparalleled. A much
less intellectual effort would be required to find out the truth for
oneself than to understand his exposition of it." l Ferrier does not
exaggerate the difficulty of historical exegesis; for it is true not
only that the great philosophies require to be distilled, but that they
also require to be translated from the terms of their own traditional
context to the terms of another. Moreover there must always be a
large marginal error in any such interpretation. This being the case,
it is not only gratuitous, but suicidal, to add the difficulties of this
problem to the difficulties of each special philosophical problem.

IV

REALISM AS A CONSTRUCTIVE PHILOSOPHY

As is almost universally the case with conscious and methodical
criticism, realism finds itself committed to certain positive beliefs.
The very act of criticism itself cannot but define, however broadly
and tentatively, the outline of a general philosophy. Thus, the
grounds on which realism rejects subjectivism determine to some
extent the superstructure which is to be reared in its place ; while
the very fact of the rejection of subjectivism excludes one of the
leading metaphysical alternatives, and gives heightened emphasis
to the alternatives that remain.

1 Ferrier. Institutes of Metaphysics, 96-97.

32 INTRODUCTION

1. Perhaps the most notable feature of a realistic philosophy is
the emancipation of metaphysics from epistemology.1 This means
that the nature of things is not to be sought primarily in the nature
of knowledge. It does not follow that a realist may not be brought
in the end to conclude that moral or spiritual principles dominate
the existent world, but only that this conclusion is not to be reached
by arguing from the priority of knowledge over its objects. Moral-
ism and spiritualism must take their chance among various hy-
potheses ; and the question of their truth is to be determined by
the place of such principles among the rest within the world. The
general fact that whatever the world be judged to be, it is at any
rate so judged, and therefore an object of cognition, is to be ig-
nored ; and one is left to decide only whether on empirical grounds
one may fairly judge the world to be spiritual or moral in part
only, or on the whole. It will be seen at once that the chief ground
on which a spiritualistic or ethical metaphysics has latterly been
urged is removed. But at the same time the metaphysical sig-
nificance of life, consciousness, and morality as facts among facts
is at once increased ; and these may now be employed for the
formulation of hypotheses that are at least pragmatic and verifiable.

2. Again, in rejecting anti-intellectualism and espousing the
analytical method, realism is committed to the rejection of all
mystical philosophies. This holds of all philosophies that rely
on immediacy for a knowledge of complexness ; of all philosophies
that regard the many in one as a mystery that can be resolved
only by an ineffable insight. A neo-realist recognizes no ultimate
immediacies nor non-relational nor indefinable entities, except
the simples in which analysis terminates. The ultimate terms of
knowledge are the terms that survive an analysis that has been
carried as far as it is possible to carry it ; and not the terms which
possess simplicity only because analysis has not been applied to
them. Such a course of procedure is fatal, not only to a mystical
universalism in which the totality of things is resolved into a mo-
ment of ecstasy, but also to those more limited mysticisms in which

i Cf. below, No. I.

REALISM AS A CONSTRUCTIVE PHILOSOPHY 33

complexes such as substance, will, activity, life, energy or power,
are regarded despite the obvious manifoldness of their characters,
as nevertheless fused and inarticulate. It follows that neo-realism
rejects all philosophies in which metaphysics is sharply divorced
from the special sciences, on the ground that while the latter must
analyze, specify, and systematize, the former may enjoy a peculiar
illumination of its own, in which the true heart of things is made
apparent, and the facts and laws of science are reduced to dead
abstractions, or mere instrumental artifacts.

3. For several reasons the new realism tends, at least in the
present state of knowledge, to be metaphysically pluralistic rather
than monistic. Most metaphysical monisms have been based on
one or the other of two grounds. The first of these is the internal-
ity of relations ; the supposition that the nature of terms contains
their relations. It is easy to argue from this premise, that since
all things are interrelated, the nature of each contains the nature
of all. Realism rejects the premise that all relations are internal,
because it is believed that it is contrary to the facts of existence,
and to the facts of logic. The second ground of monism is the
universality of cognition. The rejection of this is, as we have seen,
the very starting-point of realism. Without one or the other of
these grounds it is not possible to construct a monism dialectically
or a priori. This question also becomes an empirical question,
and in lieu of the discovery of a law, or set of postulates that shall
explain everything, we must at least remain skeptical. The evi-
dence at present available indicates that while all things may per-
haps be related, many of these relations are not constitutive or
determinative ; that is, do not enter into the explanation of the
nature or existence of their terms.

4. Again, the primary polemical contention of realism, its re-
jection of subjectivism, has its constructive implications. If
cognition is not the universal condition of being, then cognition
must take its place within being, on the same plane as space, or
number, or physical nature. Cognition, in other words, has its
genesis and its environment. When knowledge takes place, there

34 INTRODUCTION

is a knower interacting with things. The knower, furthermore,
since it cannot legitimately be saved from analysis, and referred
to a unique mystical revelation, must take its place in one mani-
fold with the things it knows. The difference between knower and
known is like the difference between bodies, or states of conscious-
ness, or societies, or colors, or any grouping of things whatsoever
in the respect that they must be brought into one field of study, and
observed in their mutual transactions.

In all this it is presupposed that if there is to be knowledge, there
must be something there to be known, and something there to
know ; ' there ' meaning the field in which their relation obtains.
Their correlation is not a basic and universal dichotomy, but only
a special type of correlation, having no greater prima fade dignity
than the many other correlations which the world exhibits. It
is not to be taken in bar» formal terms, but is to be observed con-
cretely, and in its native habitat. The realist believes that he
thus discovers that the interrelation in question is not responsible
for the characters of the thing known. In the first place being
known is something that happens to a preexisting thing. The
characters of that preexisting thing determine what happens when
it is known. Then, in the second place, when the knowing takes
place, these characters are at least for the most part undisturbed.
If they are disturbed, or modified, then the modification itself has
to be explained in terms of certain original characters, as conditions
of the modification. So that even if it proved necessary to con-
clude that illusion and hallucination are due to modifications of
the stimulus by the reacting organism, this very conclusion would
imply the preexisting and independent character of the body in
which the stimulus originated.

5. In immediate and intimate connection with this doctrine
of the independence of things known and the knowing of them,
stands another special doctrine — to the effect that the content
of knowledge, that which lies in or before the mind when knowledge
takes place, is numerically identical with the thing known.
Knowledge by intermediaries is not denied, but is made subordi-

REALISM AS A CONSTRUCTIVE PHILOSOPHY 35

nate to direct or presentative knowledge. There is no special
class of entities, qualitatively or substantively distinguished from
all other entities, as the media of knowledge. In the end all
things are known through being themselves brought directly into
that relation in which they are said to be witnessed or apprehended.
In other words, things when consciousness is had of them become
themselves contents of consciousness; and the same things thus
figure both in the so-called external world and in the manifold
which introspection reveals.

6. Finally, because he regards analysis and conception as means
of access to reality, and not as transformations or falsifications of
it, and because he asserts the independence of reality in the know-
ing of it, the neo-realist is also a Platonic realist. He accords full
ontological status to the things of thought as well as to the things
of sense, to logical entities as well as physical entities, or to sub-
sistents as well as existents.

7. In short, for realists, knowledge plays its part within an in-
dependent environment. When that environment is known it is
brought into direct relations with some variety of agency or pro-
cess, which is the knower. The knower however is homogeneous
with the environment, belonging to one cosmos with it, as does an
attracting mass, or physical organism, and may itself be known as
are the things it knows. The world is of an articulate structure
that is revealed by analysis, consisting of complexes, like bodies,
persons, and societies, as well as of simples. The simple con-
stituents of the world comprise both sensible qualities and logical
constants. Both enter into the tissue of fact, and both possess
an inherent and inalienable character of their own. There is no
safe refuge from this conclusion in any abandonment of intellec-
tual rigor. Hence all speculative versions of the world that re-
quire the withholding of analysis, or that depend on the unique
and preeminent status of the act of cognition, must be rejected,
no matter how eagerly they may be desired for the justification
of faith. They must be rejected in favor of such hypotheses as
may be formulated in terms of the evident composition of the

36 INTRODUCTION

known world, and verified by its actual interrelations, history, and
trend.

These conclusions in the aggregate can scarcely be said to be
negative. It is true that they constitute neither a complete phi-
losophy, nor, even so far as they go, an absolutely systematic phi-
losophy. But that a philosophy should be absolutely systematic
in the sense of being deducible from one principle is itself a philo-
sophical doctrine that the realist is by no means prepared to adopt.
Moreover that his philosophy should be as yet incomplete is, to the
realist at least, a wholesome incentive, rather than a ground for
uneasiness. There are endless special philosophical questions
to which there is no inevitable realistic answer, such questions as
mind and body, teleology, the good, and freedom; and there is
as yet no general realistic philosophy of life, no characteristic
verdict on the issues of religion. Nevertheless, the foundations
and the scaffolding of the realistic universe are already built ; and
it is even possible for some to live in it and feel at home.

REALISM AND THE SPECIAL SCIENCES

1. IT is the earnest hope of those who have identified themselves
with this movement, that it may afford a basis for a more profit-
able intercourse with the special sciences than that which has lat-
terly obtained. There are common problems which have been
hitherto obscured by a radical difference of method, and an in-
commensurability of terms. So long as philosophy is simply the
exploitation of a unique and supreme insight of its own, it remains
either irrelevant to the special sciences or, through its claim of
superiority, a source of irritation and an object of suspicion. Such
has, to some extent, at least, been the case during the later philo-
sophical regime. Idealists have benevolently assimilated science
to a universal consciousness ; irrationalists have appealed to revela-
tion for insight that overrules and makes naught of all the hard-

REALISM AND THE SPECIAL SCIENCES 37

won truths of science. In either case, science is not helped by
philosophy, but after being allowed to do the work of truth finding,
is graciously assigned to headquarters labeled ' Appearance ' or ' Mere
Description/ where it may enjoy the patronage of a superior.

Realism advances no all-inclusive conception under which science
as a body may be subsumed ; it claims no special revelation, and
asks no immunity from the pains of observation and analysis.
What is thus lost of eminence and authority, may, it is hoped,
be made up by a more cordial and profitable association with
fellow-workers in a common task. For, after all, the division of
the disciplines is less significant than the identity of problems and
the singleness of purpose that should animate all rigorous seekers
after knowledge. Consciousness, life, infinity, and continuity
are genuine and identical topics of investigation, whether they
happen to be alluded to by psychologists, biologists, logicians, and
mathematicians, or by philosophers. And it is reasonable to hope
that the difference of training and aptitude between the special
scientist and the philosopher should yield a summation of light,
rather than misunderstanding and confusion.

2. Thus psychology, for example, has for its very subject mat-
ter the concrete process of consciousness, and is therefore vitally
concerned in anything true which philosophy has to say about
consciousness in general. But the alleged discovery of subjectiv-
ism, that all things are mental, is so untrue to the phenomena on
which psychology has to work, that this science has been brought
thereby to a peculiar state of embarrassment. In the concrete
processes of perception and cognition, the corpus vile of psychology,
the stimuli, howsoever 'mental' they may be in some last and
remote analysis, are assuredly not mental in the sense in which the
correlated sensations and ideas are so. Precisely because the
psychologist has to accept the direct evidence for the existence of
particular minds, he can take no part in the conspiracy to make of
mind a universal predicate.

The result is that idealism has meant nothing to the actual
psychologist, who has in his laboratory remained a Cartesian dual-

38 INTRODUCTION

1st. And it is unmistakable that the results of the study of the
soul are to-day, and have been through the last three centuries,
read off and tabulated in terms of two substances — matter and
mind. Sensations and ideas, alleged to be peculiar and private
to each percipient, are conceived as invisible pawns which are cor-
related one-to-one with the ' brain-cells ' or other cerebral structures,
and are superfluous to the actual processes of the brain in spite of
frantic efforts to assign to them some regulative function; and
they have none but the most chimerical and unstatable relations
to the outer objects which these pawns are said to represent. The
supposed need of interpreting the results of empirical psychology,
or rather of 'observing' all mental processes in terms of two sub-
stances, has thoroughly stultified the science as a whole. The
artificial and unsupportable situations to which this course has
led are numerous, but one in particular is so preposterous and
unendurable that it alone would demand a complete revision of
the current 'presuppositions' of psychology. This is the concrete
situation when two persons are making a psychological experiment.
One is called the experimenter, the other the observer or 'subject,'
and between them lie the instruments for giving stimuli and re-
cording results. The experimenter, by hypothesis, has direct and
immediate knowledge of these instruments and in particular of
the stimuli which he employs. By hypothesis the observer, al-
though similarly a human being with the same gift of cognition,
has not a direct or immediate apprehension of these instruments
and stimuli, but this observer's knowledge is limited to the field
of invisible pawns which 'represent' the stimuli, and which en-
joy an otherwise inscrutable status of one-to-oneness with some
structures within the observer's skull. So the situation is inter-
preted, until presently the two experimenters exchange their rolls,
whereupon by a process of magic the just-now observer acquires
a direct apprehension of the instruments of stimulation, the scales
have fallen from his eyes and are adjusted to the other man's, whose
conscious field now shrivels and is merely the fitful flux of the in-
tracerebral and invisible pawns.

REALISM AND THE SPECIAL SCIENCES 39

This is the situation which attends every psychological experi-
ment in which two persons take part. It is absurd, and can be
mitigated only by a theory which gives a satisfactory epistemologi-
cal status to the 'outer objects' which are the terms common to
all human experiences. Neither dualism nor idealism provides
such a status. This condition of things is sufficient to induce the
psychologist to look toward realism; and yet this is merely one
of several insupportable results attendant on a dualistic psychology.
In general, it may be said that any argument which makes dualism
indefensible in philosophy makes it concretely intolerable in
psychology. Psychology has not yet found the right fundamental
categories, and will not find them as long as dualism continues to
hold sway. Meanwhile its particular findings lie accumulated in
incoordinated heaps and investigators are beginning to sense an
impasse, and are somewhat inconsequently turning away to various
forms of an ' applied ' science.

3. A similar state of things exists in biology ; for here a real-
istic philosophical basis is even more clearly presupposed. Indeed,
the realistic point of view and all its fundamental propositions may
be served on the biologist as a mandamus; for to him are assigned
such problems as the origins of life, the origins of species, the man-
ners of growth, of variation, and of adaptation. Now each and
every one of these problems presents a situation wherein there is
an environment independent of a given creature which is being
affected by that environment and is, in turn, manipulating itself
and parts of the environment. Such a world is realistic ; it is no
piece of human imagery, and its texture is made of other stuff
than mere thoughts. It is full of minds which it has somehow
made and which it, by a mere invisible lesion, can destroy.

As with the world, so with the organisms in it. They are not
the products of the minds they bear. Although these minds do not
even suspect the form and flux of their sustaining organs, yet the
latter operate, day and night, indifferent to that ignorance. They
are as independent of the mind as is the wind which sighs around
the house while the mind sleeps.

40 INTRODUCTION

It is true that many biologists look with favor upon idealistic
doctrines, which, if accepted, would lead to absurdities. They
have applied them only half-heartedly though and thereby be-
fuddled many questions, notably that of vitalism versus mechan-
ism. So far have some of them gone with the doctrine that things
are 'mental constructs,' that they have projected conscious intel-
ligence the whole organic process. But they cannot doubt that
an organism is needed to produce a 'mental state.' Thus Pauly
cannot understand how an organism could ever 'grow' eyes,
unless the cause of the growth also had eyes. Nothing, he vir-
tually argues, can be done unless the deed is known beforehand in
detail ; though to know it, the knower behind the organism must
in turn have a perceiving mechanism. It is to avoid such bewil-
derment that realism wishes to join hands with the cautious biol-
ogists.

4. If realism can afford assistance and clarification to psychology
and biology, this is no less the case with logic and the mathemati-
cal sciences. At the present time these latter sciences suffer
chiefly from a confusing admixture of psychology. This confusion
takes two forms, as illustrated by the case of logic. On the one
hand, logic as a science of such entities as terms, propositions,
propositional functions, etc., is confused with the study of the art
and processes of thought. On the other hand, logic as a science
of implication and necessity is confused with the study of the his-
torical genesis of knowledge. Realism frees logic as a study of
objective fact from all accounts of the states or operations of mind.

For the realist there are empirical grounds for holding that the
object known is independent of and may be dissimilar to the cog-
nitive process. Cognition can be eliminated. It is discovery.
Accordingly, the realist is an open-minded empiricist. He stands
quite ready to find and to admit that anything may be a fact,
that any kind of entity may exist, or subsist. The only limitations
are a posteriori. For the realist, the study of the knowing process
is only one of many fields of investigation. Logic, arithmetic, and
mathematics in general are sciences which can be pursued quite

REALISM AND THE SPECIAL SCIENCES 41

independently of the study of knowing. The entities with which
they deal are not physical ; nor are they mental. They are sub-
sistents in that they are entities notwithstanding this fact. Thus
these sciences investigate neither physical nor mental entities,
but have to do with an independent and objective field of their
own.

5. It is necessary that philosophy should raise the questions of
epistemology, if only in order to assign them a subordinate place.
It will not do to ignore the fact of knowledge itself. Sooner or
later, the knower must take himself into the account and become
conscious of that inward relation to a subjective background which,
in the first objective or outward intent of knowledge, is naturally
overlooked. Realism is not a naive or blind neglect of the prob-
lem. If realism concludes, as it does, that the knower himself
may, in the great majority of cases, be disregarded, and the object
be explained in its own terms, it is only after due consideration of
the matter. The right so to disregard the subjective conditions
of knowledge is an achievement of critical reflection.

And it is an achievement of no small moment; for it at once
establishes the full rights of all special branches of knowledge.
Philosophy must, it is true, now abandon its supposed privilege
of radically transforming all results which have been reached with-
out taking knowledge into account. Philosophy can no longer
condemn such results as necessarily and universally false, or re-
place them with a higher esoteric truth, which is revealed only to
the initiated. The disregard of epistemological considerations
which is characteristic of special investigations is now justified.

But what philosophy loses in prerogative, it gains in the improve-
ment of its relations with other branches of knowledge. It may
now employ the results of the special sciences as they stand. This
is true not only of the physical sciences, but of the moral sciences ;
and not only of scientists in the professional sense, but of all ob-
servers and investigators who have anything to report concerning
the state of things in this common world. In other words, once
subjectivism and mysticism are discredited, the work of philosophy

42 INTRODUCTION

becomes continuous with that of all who have chosen to limit
more narrowly the field of their labors. There will always remain
a certain difference of procedure between philosophers and spe-
cialists. Philosophers will be looked to for breadth of generaliza-
tion, for refinement of criticism, and for the solution of such prob-
lems as are peculiarly connected with the limits of generalization
and criticism. But, even so, the task of philosophy is not radi-
cally different from that of the special knowledges. It lies on the
same plane, or in the same field. It is a difference of degree and
not of kind ; a difference like that between experimental and theo-
retical physics, between zoology and biology, or between juris-
prudence and political science.

Thus, realism proposes that philosophy should abandon for all
time that claim to the hereditary exclusive possession of truth
which was made in the first days of its youthful arrogance. Though
philosophy has until now clung tenaciously to that dualism of
knowledge by which Parmenides assigned to philosophy "the un-
shaken heart of persuasive truth," and left for the less privileged
workers in the field of empirical facts only "the opinions of mortals
in which is no true belief at all"; it is the conviction of those
who have undertaken the present volume that the way of all
mortal opinion, in so far as it is honest and attested by evidence,
is the way of truth.

THE EMANCIPATION OF METAPHYSICS FROM
EPISTEMOLOGY

THE EMANCIPATION OF METAPHYSICS FROM
EPISTEMOLOGY

BY WALTER T. MARVIN
I

THE ISSUE BETWEEN DOGMATISM AND CRITICISM

1. THE purpose of this essay is to present some arguments
in opposition to the belief, held by many philosophers, that the
science which investigates the nature, the possibility, and the
limits of knowledge is fundamental to all other sciences and to all
other scientific procedure, and in particular that this science
either is metaphysics or is fundamental to metaphysics.1 As a
preliminary to our discussion we must clearly understand what is
meant by "one science being fundamental to another." To an
inquiry concerning the meaning of the words "one science is
fundamental to another," three answers appear to be offered.

First, one science is fundamental to another when it is logi-
cally prior; and by logical priority is meant that relation which
holds between a proposition and its necessary condition. Thus
if A implies B but B does not imply A, then B is the necessary
condition of A ; for A's truth depends upon B's truth. That is,
should B prove to be false, A must be false : and though A be
false, still B may prove true; for we are saying merely that A's
truth is a sufficient condition of B's truth, and are not maintain-
ing that it is the only condition, or a necessary condition. For
example, let us assume it to be true that if the tissues of a man's

1 Under the term metaphysics I include two subjects : (a) the study of the
logical foundations of science ; (6) the theory of reality. Here and throughout
this essay I mean by the words, theory of reality, any collection of fundamental
existential propositions and of high existential generalizations.

45

46 METAPHYSICS AND EPISTEMOLOGY

body absorb a certain amount of arsenic, he must die, that there
is no preventing cause either known or unknown. Then evi-
dently for it to be true that this man's body has absorbed such
an amount of arsenic, it must be true that the man is dead ; whereas
the mere fact of his death does not prove that many another pos-
sible cause is not the actual cause. In short, "the man is dead"
is logically prior to the proposition, the tissues of his body have
absorbed the required quantity of arsenic. But let us illustrate
specifically the logical priority of one science to another. Much
of mathematics is logically prior to mechanics and physics, since
much of these latter sciences could prove false without thereby
indicating any errors in our pure mathematical theories; but
should it be found that arithmetic, the calculus, and elemen-
tary geometry are false, evidently our mechanical and physical
theories, based as they are upon these sciences, and being in great
part explicit deductions from them, would fall to the ground.
Of course there may be other, as yet totally unknown, ways by
which mechanics and physics can be demonstrated; but accord-
ing to our present knowledge, unless a large part of mathematics
is true, mechanics and physics must be false. If then we accept
the foregoing as the meaning of the word 'fundamental/ we get
as the first answer to our question : In calling the theory of knowl-
edge fundamental, the philosopher means that it is logically
prior to all other knowledge.

2. The second answer offered appears to be different, though
a closer scrutiny may reveal the presence of the same conviction.
"The theory of knowledge is fundamental," means not only that
the epistemologist can ascertain through his science the limits
of possible knowledge, but especially that he can do so without
studying the various special sciences or the history of science and
of scientific discovery, or without in any way going for informa-
tion beyond the territory of his own science. In other words,
it is maintained that there is a science of the possibility of knowl-
edge which is not an induction from men's scientific experience
during the centuries of civilization nor from the sciences as they

ISSUE BETWEEN DOGMATISM AND CRITICISM 47

are now known, but which is a direct, independent, and final
study of the nature and limits of knowledge. Indeed, it is held
that unless we have proven in this way the possibility of any
special science, such a science is mere dogmatism ; and therefore
that he only is a critical scientist who does not attempt to inves-
tigate until he proves to himself by the theory of knowledge that
what he hopes to discover and to explain is a possible object
of knowledge.

Of course the writer has in mind especially Kant and his ' Cri-
tique of Pure Reason.' He it was who taught, as no other man has
taught, that dogmatism and criticism are forever irreconcilable,
that one science, the science of the possibility of knowledge, can
ascertain what are and what are not problems which the human
mind can solve. He thought that the older metaphysicians,
who were either ignorant or careless of these matters, were led
hopelessly into error precisely because they undertook to solve
their problems before they wrote or studied a critique of pure
reason.

Moreover, note well; this critique of pure reason is not a his-
tory of the successes and failures of scientists, nor is it a summary of
what in the course of human history have proved solvable and
insolvable problems. Rather it is a direct study of the nature of
knowledge in the abstract and of the behavior of the human in-
tellect; and its conclusions are said to be drawn from this study
without the aid of the other sciences. On this account its results
are believed to be independent of the special sciences and au-
thoritative over them ; whereas if it drew its information from them
or were itself a mere induction from human experience, it would
be admittedly a vicious circle.

3. The third answer informs us that in calling the theory of
knowledge fundamental the philosopher asserts that this science
can enable us to ascertain the validity of the special sciences
and of their methods, and in certain respects at least can enable
us even to correct their results. For example, if a science offers
us a solution of some problem which we know to be insolvable,

48 METAPHYSICS AND EPISTEMOLOGY

or asserts what is beyond the possibility of human knowledge,
we can infer at once that this part of such a science must be in-
valid. But this is by no means all that is referred to in the third
meaning of the word, 'fundamental.' Rather we are given to
understand that the theory of knowledge offers us information
regarding reality, and that this information is of great value in
two respects ; it is more certainly true than are the results of the
special sciences, and it is a means of refuting, correcting, and
limiting these results. For example, if a study of the nature of
knowledge shows that the universe must be an organic unity;
it can be inferred that should the sciences indicate the opposite,
they are at the best only relatively or partly true. Again, if a
study of knowledge shows that only mental contents can be known,
that an object to be known must be part of the mind's experi-
ence; then it can be inferred directly without further evidence
from science that reality as far as it is knowable must be the
experience of some mind. Then if in addition the theory of knowl-
edge shows that whatever is essentially unknowable cannot be
real; we reach the conclusion, reality as such is the experience of
one or more minds, and only that. Hence, should popular opinion
and scientific inference assert in opposition to this that things
exist which cannot be experienced or which belong to no one's
experience, such doctrines would be subject to correction by our
fundamental science.

Still again, if it be true, as some epistemologists have taught,
that the mind in knowing gives a form to the objects known and
consequently that whatever is known must have this form or
structure; then their science can lay down for all time to come
the main outlines of the world as the possible object of scientific
research. Should the physicist or any other scientist ever quar-
rel with this outline furnished by our philosopher, it will be our
duty to inquire whether or not he is a student of the theory of
knowledge and has made a discovery in that field. Should it
prove that he is not an epistemologist, but just a physicist, mathe-
matician, or chemist; our philosopher will tell him that he is not

ISSUE BETWEEN DOGMATISM AND CRITICISM 49

competent to talk about ultimate reality, for his science may
indeed give useful information, but cannot give any independent
and fundamental insight into what is and what is not ultimately
real. For example, it has been claimed by one epistemologist
or another that the world as an object of knowledge must be a
three-dimensional spatial system, a temporal system, a causal
system, a system of sense impressions, a society of minds, an
organic unity, an infinite and perfect personal mind, a divine
language by which God sends us messages, a battlefield created
by the mind wherein the will may fight for the moral ideals, the
dreary, hopeless outcome of the struggles of an impersonal, blind
and restless will, an evolution of an absolute mind. Negatively
it has been held by one philosopher or another as the result of
his study of the nature of knowledge, that matter does not truly
exist, that colors and sounds, heat and cold, do not exist outside
the mind, that scientific laws are not truly parts of nature, that
the real world cannot be known.

4. Let us sum up briefly these three meanings of the state-
ment, "the theory of knowledge is fundamental," in the following
propositions: First, the theory of knowledge is logically prior
to all other knowledge; secondly, one can by a direct study of
the knowing process infer the limits of possible knowledge; and
thirdly, the student of epistemology can give us, independently
of all other sciences, a theory of reality. By no means do I claim
that these three propositions cannot be and should not be further
analyzed and reduced to one, namely to the first ; rather, I believe
precisely this: but for the purpose of the present argument it is
better to leave them as they stand.

5. In opposition to these beliefs in the fundamental charac-
ter of the theory of knowledge, this essay will support directly or
indirectly the truth of the following propositions : (a) first, that the
theory of knowledge is not logically fundamental, that on the con-
trary its logical position is posterior to many of the special sciences,
such as physics and biology ; (&) secondly, that the theory of
knowledge does not enable us to show, except inductively and em-

50 METAPHYSICS AND EPISTEMOLOGY

pirically, either what knowledge is possible or how it is possible,
or again, what are the limits of human knowledge; (c) thirdly,
that no light is thrown by the theory of knowledge upon the na-
ture of the existent world or upon the fundamental postulates and
generalizations of science, except in as far as the knowledge of one
natural event or object enables us sometimes to draw inferences
regarding certain others; (d) fourthly, that epistemology does
not give us a theory of reality, on the contrary, -it assumes one ;
(e) finally, that it neither solves metaphysical problems nor is it
the chief source of such problems. One may express all of this
affirmatively as follows. I shall try to show three things: (a)
first, that the theory of knowledge is one of the special sciences,
that it studies knowledge as a natural event and in virtually the
same way and by the same methods as biology studies life or
physics light ; (6) secondly, that as such a science, it assumes the
formulae of logic and the results of several special sciences, such
as physics and biology; (c) and finally, that logic, metaphysics,
and some existential sciences are logically prior to the theory of
knowledge.

6. In short, the general conclusion which I shall draw is that
metaphysics is logically prior to the theory of knowledge and
that it is not peculiarly indebted to this science either for its
problems or for their solution. If this conclusion is true, then
metaphysics should be completely emancipated from epistemology ;
for the sway that this science has held over metaphysics from the
days of Locke to our own time is a thoroughly unconstitutional
assumption of authority. Thus in a certain respect I am urging
a return to the old days of the seventeenth century, to the days
of Descartes, Spinoza, and Leibniz, to the method that Kant con-
demned as dogmatism. Indeed, let us for the sake of brevity ac-
cept throughout this essay Kant's terms to indicate the two oppos-
ing tendencies. In a narrow and technical meaning of the words,
the one tendency is dogmatic and its doctrines are dogmatism,1

1 It should be distinctly understood by the reader that the word dogmatism
is used throughout this essay in the narrow and precise sense above denned. The

EPISTEMOLOGY NOT FUNDAMENTAL 51

whereas the teachings of the opposed tendency are criticism
and their defenders are criticists.

II

THE THEORY OF KNOWLEDGE IS NOT LOGICALLY FUNDA-
MENTAL

1. THE first and most prominent tenet of the criticist may
be stated thus : Inasmuch as all sciences are cases of knowl-
edge, the science which investigates knowledge as such is funda-
mental and is, both in fact and by right, a critique of all science.
Underlying this doctrine the dogmatist finds, or at least suspects,

name is taken from Kant's Critique of Pure Reason where, whatever else it may
mean, it denotes the contradictory of what Kant calls criticism. Unfortunately,
the word has other associations in Kant's mind and in the mind of the student of
Kant ; for it sometimes means specifically the rationalistic ontology of the Cartesian
and Leibrdzian philosophers, whereas neo-realism differs radically from this phi-
losophy. For example, many neo-realists have a strong tendency toward an extreme
empiricism and toward an abandonment of the substance-attribute notion as a
fundamental notion in metaphysics. Again, neo-realism is epistemological monism ;
whereas the Cartesians were epistemological dualists, holding to a representative
rather than a presentative theory of perception. Finally, a modern dogmatism
must of necessity differ from that of the earlier centuries just because it has behind
it two centuries of experience with criticism. That is, it is consciously and de-
liberately dogmatic, whereas the earlier dogmatism was naive and was therefore
easily misled into idealism and its so-called criticism. But in spite of these unfortu-
nate associations I believe the names dogmatism and criticism not only appropriate
but enlightening ; for I think the neo-realistic movement to be a reaction against
the whole enterprise of Locke, Kant, and their followers to get at a fundamental
science and not merely against their idealism. That is, neo-realism is not only a
different theory of knowledge but, what is more important for metaphysics, a
different doctrine as to the place of epistemology in the hierarchy of the sciences.
As the names realism and idealism do not point out this difference clearly; I prefer
the names dogmatism and criticism, which, if taken in their generic meanings as
given by Kant, certainly indicate precisely this difference. Indeed, I would go
further ; for many contemporary realists are criticists, and it is at least conceivable,
no matter how remarkable, that some dogmatists may be idealists. My points
may be summed up briefly in the following two sentences. Dogmatism is the con-
tradictory of criticism and defines neo-realism negatively or by exclusion. Chiefly
and perhaps only in this respect is neo-realism a return to seventeenth century
philosophy.

52 METAPHYSICS AND EPISTEMOLOGY

two errors, on the one hand the assumption of a false theory re-
garding the nature of logic, and on the other a failure to distin-
guish between two uses of the word 'knowledge,' that which
denotes the act of knowing and that which refers to the truths
or propositions known.

2. To many philosophers logic still seems to be a science of
the knowing process, or more precisely, a science of the laws of
thought, that is, of the rules dictated by the mind's own nature
and obeyed by us whenever we think correctly; whereas logic
is nothing of the sort. The formulae of logic are no more laws of
thought than is the undulatory theory of light, the Mendelian
law of heredity, or for that matter a recipe for cake or even an
adding machine. Logic gives us no information in particular
regarding the mind or the thinking process; and the logician's
views on such subjects might be quite erroneous without leading
him astray within his proper field.

3. What then is logic ? And I mean by logic not only the re-
sults of recent study, which the reader may or may not value highly,
but also the ancient doctrines to be found in the writings of Aris-
totle and in the textbooks of past centuries. The logician offers
us, as does any other scientist, information regarding certain terms
and their relations. Some of these terms are classes, and some
of these relations are the relations obtaining between classes and
their members or between one class and other classes. Further,
some terms studied in logic are propositions, and propositions
are found to be related in a way called implication. Therefore
the logician tries to learn the ways in which one proposition can
be related by implication to another. Finally, logic deals with a
number of fundamentally different sorts of relation. As the logi-
cian puts it, some are transitive, some intransitive, some sym-
metrical, some asymmetrical, and so on.

4. Now in all of this logic is studying something non-mental
in the same sense as does mathematics or chemistry. There are
in the world about us classes and these classes are related. There
are such things as truths and falsehoods and these as such are

EPISTEMOLOGY NOT FUNDAMENTAL 53

related to one another. Moreover, they are so related quite
apart from any question of human existence or human thought.
" Two plus two equals fourteen," was false fifty million years ago ;
and the fact that it was false, made the world of that day a very
different world from what it would otherwise have been. Thus to
the best of our knowledge, the fact that one proposition implies
another is not merely a pleasant and playful thought of this man
or that, but it is a downright serious matter. It seems to deter-
mine what happens in this world about us, it seems to determine
whether a man shall die or live, shall be born or not be born,
shall be happy or utterly wretched. It seems to determine even
whether a solar system shall go along peacefully and evolve habit-
able planets, or shall go to smash and end in chaos. There may
be some sense in which all of these things are mental; that is,
some sense in which astronomy, physics, chemistry, biology,
geology, and what not other science is a study of human knowl-
edge and of the knowing process. All well and good; if such be
the case, no doubt logic is so too; but if in other respects they
are not such a study, then neither is logic. The nature of the
physical universe depends upon whether or not logic is true as
genuinely as it does upon the truth of this or that physical theory.
Therefore the logician has a right to say: "When I study classes
and their relations, or propositions and their relations, I am
studying aspects of the world about me as truly as does the physi-
cist when he studies the nature of light, heat, gravity, and elec-
tricity."

5. "But," you ask, "is not logic the science or art of correct
reasoning? And is not reasoning a mental process?" No, logic
is not. Of course there is such a study or art, and of course there
is excellent authority for the use of the word logic as the name
of this art. But the art called logic, when examined critically
from the point of view of the pure sciences, is a conglomerate of
many sciences applied to solving one type of practical problem.
In short, it is the application of information from many scien-
tific sources. It draws on pure logic, it draws on psychology, it

54 METAPHYSICS AND EPISTEMOLOGY

draws on mathematics; indeed, I decline to mention the pure
science upon which it should not draw.

6. Yet it may be protested, "In all our reasoning we use logic,
therefore logic is the science of reasoning." Such an argument
is fallacious, and what is more, its conclusion is false. To make a
long story short, we must define what is meant in this argument
by the word 'use,' and we must decline from the start to reckon
by percentage of use whether or not this or that science or part of
a science is a study of the knowing process. Do we in our reason-
ing use logic in a different way from that in which we use mathe-
matics, physics, chemistry, or astronomy? Now if we do not,
and if the only difference is that we use some parts of logic every
time we reason, why should we then draw the line at one hundred
per cent (really a lesser per cent for parts of logic may be used
quite infrequently and we should take the average) and not at
forty-five per cent? Evidently the man who believes logic to
be a science of reasoning is not thinking of percentage of use.
Rather he holds that the use our reasoning makes of logic is differ-
ent from the use our minds make of mathematics or chemistry.
It is then all a question of the meaning of the word use.

How do we use logic in our reasoning? I reply, in the same
way in which we use physics. How is that? We make use of
the laws or propositions of physics as premises or as formulae for
whose variables we substitute constants. Let me illustrate. I want
to know how far a projectile will go if it leaves the ground at a
given angle and at a given velocity. Physics gives me formulae
from which, if I use as premises along with the given conditions also
used as premises, I can infer the proposition which I wish to know.
Again, mathematics tells me (a + ft)2 = a2 + 2 ab + b2. I want
to know the square of 27. How then do I use (in my reasoning)
this information? We substitute, let us say, for a 20 and for
b 7; that is, we substitute constants for the variables in the equa-
tion. Thus (20 + 7)2 = 400 + 280 + 49 = 729. Hence to use
physics or any other exact or natural science in our reasoning is
to adopt its propositions as premises, Now is the same thing

EPISTEMOLOGY NOT FUNDAMENTAL 55

true when we use logic in our reasoning? It is. The results or
truths of logic are assertions, as we have said, regarding the re-
lations of classes and propositions. Further, these results of
logic are usually formulse, that is, propositions whose terms are
variables. Thus, roughly stated, if any class a is contained in
another class b, and if this class 6 in turn is contained in a third
class c, then the first class a is contained in the third class c ; or
more precisely stated, [(a <b)(b< c)] implies (a < c), where a, 6, c,
represent any class. Here is a logical formula taken from the
logic of classes. How do we use it in our reasoning? Assuming
it as true, we substitute constants for its variables. For example,
if this formula is true, and if the class men is included in the class
mortals, and if Socrates is a member of the class men, then Soc-
rates is a member of the class mortals. Every student will agree
that logic is not concerned with Socrates or man but with some-
thing more general. But notice what this means: Logic is con-
cerned with variables. It gives us formulae. If so, and if we use
logic always in our reasoning, we shall find, no matter what in-
stance of reasoning we may take instead of the trite example afore-
given, that some formula is presupposed by it. That a formula
is presupposed means that it is assumed as a premise which
we have used by substituting constants in place of its varia-
bles. In short, to use logic means to substitute in a formula con-
stants for the variables of the formula and then to assert one of
the resulting propositions, namely, the one found in that part of
the formula called the conclusion. But this, we know, is pre-
cisely what we do when we use physical formulse in our reasoning.
7. The mere fact that logical formulae are used so widely, the
mere fact that physics itself presupposes parts of logic, does not
alter the essential nature of the use. It would be perhaps a more
serious matter to have logic false than to have the undulatory
theory of light false ; but of what true proposition can we not say
something similar? Practical importance then does not deter-
mine whether or not a science is actually a study of beasts, rocks,
stars, or ocean currents ; for that depends upon the terms to be

56 METAPHYSICS AND EPISTEMOLOGY

found in the propositions constituting the science. So too in the
case of logic, whether or not logic is a study of the knowing process
depends upon what terms are found in its propositions. Now these
terms are as non-mental as are rocks and ocean currents.1 Hence,
one must draw the general conclusion : Logic is not a science of
the knowing process. Its principles and formulae are not laws
of thought. Its terms and relations are as clearly distinct from
those of thought as are the terms and relations of physics.

8. It has been stated above that a second error also is believed by
the dogmatist to underly the criticist's assertion that the science
which investigates knowledge as such is fundamental, to wit, the
criticist fails to distinguish between two uses of the word ' knowl-
edge ' : first, that which denotes the act of knowing, the natural
event called knowing, or 'knowing in the making' ; and secondly,
that which refers to the truths or propositions known, the systems
of propositions called, for example, the sciences. As a consequence
of this neglect, the dogmatist believes, the criticist ascribes to
sciences, for instance to mathematics, that which is true only
of the student of mathematics, as such a student learns, thinks,
or makes discoveries in the course of his mathematical research.

Hundreds of things may be true of this or that mathematician at
work studying his science, which are not true of mathematics.
He may be dependent upon visual pictures in his geometrical
research. He may make an important discovery by mere acci-
dent or a happy and brilliant intuition. He may be more suc-
cessful studying while he smokes than when he is not smoking.
He may be more inventive and mentally acute mornings and

1 Of course any such discussion can be a dispute about mere words, but we deny
that this is true of the foregoing. Logic is an ancient science, and it is possible,
without any idle disputation, to ascertain what its nature is. Now, as is the case in
the human history of almost every science, the student has not always seen clearly
what the nature of his science is. Thus we are far better able to-day to define logic
than was Aristotle in his time, precisely as we are better able to define geometry
than was Euclid. If to this the reply is made, " Any attempt to define a science is
either a purely verbal matter or an idle matter," I can but reply, it seems to me just
the opposite. To be able to define a science rigorously and correctly is to pass a
most important and significant milestone in the course of human history.

EPISTEMOLOGY NOT FUNDAMENTAL 57

nights than he is afternoons. In short, health, fatigue, fresh air,
digestion, season of the year, time of day, inborn mental and physi-
cal traits, previous training, praise and fame, example and com-
petition, ideals and curiosity, and what not else may all be fac-
tors in determining what he learns, what he discovers, and the
rigor of his demonstration. But what has all this to do with
mathematics? Is it part of mathematics? Would any sane
man put into a rigorous mathematical demonstration memoranda
regarding his health, the time of day, and the state of the weather,
his ambitions, and his mental imagery ? Yet why not ? Clearly
because such information is not mathematics. True, he might
state in a book on geometry the date when a proposition was dis-
covered and first proved, and who the discoverer was: but if
he did so, it would not add one whit to the mathematical infor-
mation he was giving the reader; and if he failed to do so, it would
not lessen either the accuracy of the geometrical doctrine or the
rigor of its demonstration.

9. In other words, propositions, and they alone constitute a
science, are not events in time. They do not come into being
or get created by the student who first learns that they are true.
They are discovered and not made, as truly as was the American
continent discovered and not made by the explorers of the fif-
teenth and sixteenth centuries. Thus mathematics as a system
of true propositions has been in part discovered by man; but
this discovery or that failure to discover did not add or sub-
tract anything to or from mathematics, did not make any of
its propositions either true or false, did not alter it in any way.
Two plus two equaled four, and the sum of the angles of a plane
triangle in Euclidian space equaled two right angles when the
earth was a molten mass, as truly as they do to-day. Mathe-
matics and any other science is what it is for only two reasons :
because certain propositions are true and others false, and because
one proposition implies certain propositions and does not imply
certain others. Thus if a physicist is asked why it is true that the
cables of a suspension bridge in hanging from tower to tower

58 METAPHYSICS AND EPISTEMOLOGY

formed a catenary before the rest of the bridge was constructed,
and now when the bridge is completed form a parabola, he will
not talk in reply about the knowing process, but he will show
that these propositions are true because other propositions are
true (that is, certain propositions in elementary mathematics and
mechanics) and because these propositions imply the state of affairs
mentioned (that is, the propositions asserted) in the question.1
At this the nominalistic reader may rise in scorn and protest:
"You are confusing mere abstractions with real things. You
are taking the words of men and treating them as timeless super-
natural entities, dwelling in a Platonic world of pure thought.
Apart from the thoughts of men, mathematics, or any other sci-
ence, has no more existence than has the man in the moon." To
this I reply : I have no desire either to refute or to support nomi-
nalism in what I have said ; but I do wish the nominalist to under-
stand me, and I fear that up to this point he may not have done
so. He certainly has not if the words which I have just put in
his mouth are truly there. Of course mathematics as a timeless
system of true propositions does not exist in the sense that the
Rocky Mountains or the Atlantic Ocean exists ; but on the other
hand, when man discovers a mathematical truth, he truly discovers
it as he truly discovers and does not create the distant islands
of the South Seas. Hence, whatever may be the full sense of the
statement, mathematics has its being apart from man's thought, it
will include or imply two propositions at least: first, mathematics

1 All this does not mean that the word ' science ' has not the same ambiguity as
has the word ' knowledge.' Quite the contrary, the word ' science ' means two
distinct things : on the one hand, a part of human achievement, a thing that has
had a growth and a history, a thing that can prosper or perish, a thing that de-
pends upon man for its existence ; on the other hand, a collection of propositions
that do not exist in time, that are discovered by man but in no way are made or
altered by this discovery, that would subsist and would be true or false even had
man never existed. Such is usage, and it would be in vain as well as unnecessary
to attempt to avoid this custom. Hence throughout this essay I shall use the
word 'science' in both senses, speaking for example of the history and growth of
science as well as of the propositions of science. Which is the intended meaning
in each case the context should always make quite clear.

EPISTEMOLOGY NOT FUNDAMENTAL 59

does not depend for its truth upon our minds any more or in any
other sense than does the existence of the North Pole ; and secondly,
the historical origin of man's knowledge of mathematics is in no
sense mathematical, nor does mathematics presuppose logically
any propositions regarding man, his knowing process, or in gen-
eral any proposition constituting a theory of knowledge. If mathe-
matics is true, it is so for the same reason that any other pi$posi-
tion in all reality is true, which seems to mean the truism, it is
true because reality is just what reality is.1 Again, if mathematics
is true to-day, it always was true and always will be true, because
we mean by truth something in no way a function of the time at
which it is asserted by the knower. Mathematics, then, or any
other science, does indeed not exist ; but it is, has being, subsists,
and as such it is a timeless system of propositions.2

1 That is, the question why a proposition is true can mean one of two things,
the first of which admits of an answer and the second does not. A proposition is
true because some other proposition is true and implies it. But why is that other
proposition true, why ultimately is anything true that is true ? Well, the question
is as absurd as the question, Why is red red? The question asks us to go beyond
the ultimate, and its absurdity shows us that truth is ultimate and as such is only
to be discovered, and is not to be ascertained by any device which would make it
explicable.

2 1 hope that the pragmatist reader also will not misunderstand the foregoing
statements. Of course, knowing is a natural event and as such its nature is to
be ascertained by a frank, unprejudiced study of fact and not by any dialectic.
Or, as I should prefer to put it, logic throws no light upon the nature of the knowing
process except in the sense in which it is true that logic throws light also upon the
nature of the rocks, the ocean currents, or anything else that exists or takes place.
Therefore, in as far as the pragmatist is against that type of epistemology which is
chiefly dialectic, I am heartily in sympathy with him. But as he is liable to suspect
anything which sounds like dialectic, he will no doubt think the foregoing statements
at best disguised error. If so, I believe either he fails to understand me or he mis-
takes quite the nature of logic. If he rightly permits the mathematician to go ahead
with his mathematics undisturbed, why should he interfere with the formal logician ?
The whole preceding problem is strictly and solely one of logical analysis. Formal
logic defines the relation logical priority, and holding to that definition I have
urged that the theory of knowledge is not logically prior to logic or to physics.
Questions of fact are in no way involved. If they were, of course the foregoing argu-
ment begs the question. If the problem is solely one of logical analysis, as I claim
it to be, then either there is no such rightful procedure as logical analysis, or I must

60 METAPHYSICS AND EPISTEMOLOGY

10. If then logic is not a science of the knowing process, and if
most of the sciences do not presuppose any information regard-
ing the knowing process, what is the place of the science of knowl-
edge relatively to that of other sciences? This question will
be answered more fully later on; but for the present it suffices
to say, the knowing process, the act of discovery, man's reasoning
and the conditions of this reasoning, are natural events. They
take place at definable moments; and in all essential respects
they are like other natural events which lend themselves to our
study and research. As a consequence, the science of knowledge
instead of being sui generis, and instead of occupying an excep-
tional position relatively to the other sciences, is simply one of
the special sciences. Its implications may or may not be wide
reaching, a matter to be ascertained only by a study of the facts
and certainly not by any a priori consideration of the field of the
science.

Ill

THE LOGICAL POSITION, RELATIVELY TO THE OTHER SCIENCES
OF THE THEORY OF KNOWLEDGE AND, IN PARTICULAR, OF
THE PROBLEM OF THE POSSIBILITY OF KNOWLEDGE

1. AGAINST all of the foregoing statements some criti cists will
immediately urge : "Logic, as any other science, has to assume the
possibility of knowing and, in particular, the possibility of know-
ing matters logical. Hence, there must be a science prior to all
others, even to logic, which shows the possibility of knowing.
Or if logic is indeed fundamental and therefore has to be ex-
cepted from this axiomatic statement, then it alone is prior to
the science of the possibility of knowledge, but all other sciences
are subsequent and dependent."

be permitted to pursue my enterprise undisturbed as long as I am not surreptitiously
introducing information unattainable by logical analysis, and as long as I keep to
information so commonplace and so much a matter of course that to prove it true
by a fresh investigation of fact would be needless.

THE LOGICAL POSITION OF EPISTEMOLOGY 61

Let us examine the extreme position first, "even logic presup-
poses the conclusions of this ultimate science, the science whose
subject matter is the possibility of knowledge." How is the
criticist to avoid here a vicious circle? It will be his business
to show that knowledge is possible and to show the conditions that
make knowledge possible, yet in doing this he too will have to use
premises and among these will be one asserting that his investiga-
tion as an instance of knowing is possible. Now if it is permissible
for the criticist to make this assumption in the pursuit of Ms re-
search, why may it not be permissible for the logician to do the
same in his study, and similarly any other scientist? In other
words, if it is the business of Kant in his ' Critique of Pure Reason '
to show how mathematics is possible, whose business is it to show
how the ' Critique of Pure Reason ' itself is possible ? Moreover,
if there were such an ultimate science, it would presuppose parts
of logic in the course of its demonstrations, and therefore the
criticist would be in the uncomfortable position of assuming the
possibility of logic in order to prove the possibility of logic.

2. The extreme position then involves a vicious circle and is
untenable; but is not the other position also untenable? That
is, if the logician and the criticist may rightly assume the possi-
bility of discovering and demonstrating the propositions of their
sciences, for what possible reason do they forbid the mathema-
tician, the physicist, the biologist, the historian, or any one else
from doing likewise in other branches of scientific or popular
research? Do they do so by showing that it is less difficult to
read and understand Kant's ' Critique ' than to learn how to light
a fire or to shoot an arrow? If so, our savage ancestors must
have had remarkable intellects. Do they do so by showing that the
mathematician and the astronomer had to wait until the eighteenth
century of our era to get their doctrines well established ? If so,
history proves the contrary. Do they do so by showing that
the premises upon which logic and criticism rest are self-evident,
and that their doctrine is infallible? Even Kant admits the in-
fallibility, nay, explicitly bases his views upon the assumed in-

62 METAPHYSICS AND EPISTEMOLOGY

fallibility, of mathematics and mechanics. Do they do so by
showing that the science of the possibility of knowledge presup-
poses only the results of logical research and not the results of any
other science ? That is, do they show that they are not guilty of
a similar vicious circle here in respect to some of the other sciences,
as they were shown to be in respect to logic; for perhaps they
presuppose the possibility of certain sciences in order to prove
the possibility of these same sciences ? Indeed, I shall give rea-
sons later for thinking that they do precisely this. Finally, if
none of these implied objections be true; do they show that we are
so much better able to observe directly and accurately the facts
involved in knowing that there is no need to have a science of
the possibility of knowing how we know and how we can know,
whereas it is necessary to have such a science to show that we can
know the trees, the birds, the rocks, the earth, and the stars?
Some criticists no doubt would try to show this; but in the se-
quel I shall endeavor to prove that their theory also assumes the
possibility of observing precisely these out-of-door things and so
make evident that they also assume the possibility of such knowl-
edge in order to demonstrate its possibility.

3. To put it affirmatively, I am convinced that either the
possibility of knowledge is not the premise of any science, or it is
the premise of all sciences and the conclusion of none; and of
these alternatives I believe that the former alone is true. I ar-
gue thus: The word 'possibility' contains the ambiguity pre-
viously referred to ; that is, either it means the possibility of man's
discovering and demonstrating science, or it means logical possi-
bility. The former, which seems to be the meaning in the mind
of the criticist, lies, as we have seen, entirely without the various
sciences, and has nothing to do either with their content or with
their validity. Mathematics certainly has no premise, "mathe-
matics is possible knowledge." On the other hand, if logical pos-
sibility be meant, it can refer only to the question, whether or not
the premises from which the results of a given science follow are
true. In other words, what a science assumes is not the possibility

THE LOGICAL POSITION OF EPISTEMOLOGY 63

of its being known, but the truth of its premises; and, this is even
more to the point, what you and I assume in believing this or that
doctrine to be true, for example, in believing that light is due to an
undulating motion in the ether, is not the possibility of knowing
such matters but again the truth of the premises upon which our
particular demonstration rests. Hence the statement, "it is im-
possible for us to know such propositions," could mean only, "we
cannot know them to be true " ; but if we cannot know them to
be true, this must be because we cannot know that the premises
are true, from which they follow. In short, in believing the
sciences we assume their premises to be true, and this is our only
assumption.

4. But, you retort, "Ought we not to ascertain whether or not
these premises are true?" Yes, by all means; but to assume
the possibility or even to know the possibility of our doing so,
will not help us actually to get the information : for either we
can get it or we cannot ; and if we can, we do, and if we cannot,
we do not. "Yes, but there is a further question which remains
entirely unanswered," you may reply. "In accepting these
premises are we not able to state which ones are merely assumed
to be true and which ones are true ? If so, then we must know
when we are merely assuming and when we really know. But
how can we tell the difference unless there be some ultimate
science which gives us infallible criteria by which ignorance can
be distinguished from knowledge and by which the field of possi-
ble knowledge is marked off forever from that of impossible knowl-
edge? Of course you can take the position that we always as-
sume and never genuinely know ; l but, if you do, why do you
believe some things and not others ? Why do you not believe every-
thing? Is your choice purely whim? If it is not, there must
be some infallible criterion to guide you at least in making as-
sumptions; even the pragmatist seems to have that much."

Well, some dogmatists admit the force of this argument, but they

1 I confess that some dogmatists seem to take this position ; and if they do, I
believe that in their case the criticist has the better of the argument.

64 METAPHYSICS AND EPISTEMOLOGY

draw a quite different conclusion. Let us grant that a thorough
logical analysis of any man's knowledge shows the premises of
that knowledge to be divided into two classes; first, premises
that are merely assumed and therefore are tentative, and, secondly,
premises that are not tentative, but are out-and-out fact. Other-
wise expressed, let us admit that some premises are assumed to
be true, and that some are known (or perceived) to be true.1 What
follows ? Does it follow that we cannot know a proposition to be
true without assuming the possibility of this knowledge, or, in other
words, without assuming that we know that we know it to be true ?
If so, we have an infinite regress on our hands ; and it would be
far better frankly to admit that we do not know anything, but as-
sume everything we assert.2 But this does not follow. If we
do indeed perceive, or know some propositions to be true, then
this knowledge is ultimate, and no further assumptions, premises,
or explanations lie logically behind it. If you still persist in ask-
ing how it is possible to perceive a truth, I have to reply, your
question is as absurd as the questions : Why is hard, hard ? Why
is blue, blue? I perceive it, and that ends the matter. More-
over, this must be the criticist's own real position ; for either his
epistemology, as an ultimate science, is mere assumption, or it
is based at least in part upon perceived truth. If the latter, he
has either to admit that this calls for further investigation, or to
affirm with me that it is ultimate and no investigation can go
logically behind it.

Even many realists may think that I go too far in asserting this
extreme dogmatism. They may urge: " Unless knowledge is pos-
sible, you cannot know, that is, knowing presupposes the possibil-
ity of knowing. Again, if as a realist you maintain that we
perceive the physical world truly as an extra-mental world, you

1 Cf. Stout, G. F. Immediacy, Mediacy, and Coherence. Mind, 1908, N. S.,
17, 20.

2 As far as I can see, this is the position of those who hold to the organic or monis-
tic theory of truth. Cf. Russell, B. Philosophical Essays, " The Monistic Theory
of Truth " ; also Stout, loc. cti.

THE LOGICAL POSITION OF EPISTEMOLOGY 65

thereby presuppose some theory of perception, which, if true,
would make such an act possible. In short, unless realism is true,
we do not perceive an extra-mental world ; or, our perceiving an
extra-mental world presupposes realism." Such opinions seem to
me to indicate fallacious reasoning and to be an invalid objection
to my 'extreme' dogmatism. Knowing as an event has indeed its
necessary conditions. Therefore when I assert that knowing is tak-
ing place I do presuppose all the necessary conditions of knowing, pre-
cisely as when I assert that water is boiling I assert implicitly all the
necessary conditions of water boiling. Of course, but all of this is
beside the issue. The issue is this : when I assert that water is boil-
ing, am I ipso facto asserting that I am knowing that water is boiling ?
No, for "water is boiling" and "I perceive that water is boiling"
are two different propositions and have different presuppositions.
The former proposition has no presuppositions, as far as physical
science informs me, regarding the knowing process or its possibility ;
whereas the latter proposition has such presuppositions. Now I
urge that if this last statement were not true, the science of the pos-
sibility of knowing would be an unavoidable vicious circle. As
any other science, it too has its presuppositions, and has, I sup-
pose, crucial tests for its various theories. Hence the epistemologist
is no better off than is the chemist, for as the chemist he must assume
postulates and observe facts in order to discover and to demonstrate
his theory of the possibility of knowledge. If in so doing he as-
sumes the possibility of knowing (as he claims the chemist does),
then he assumes the possibility of knowing in order to prove that
knowing is possible. In short, we have logically to start some-
where, and I maintain that it is a matter verifiable by ordinary
empirical study that the scientific investigator does not start with
propositions regarding the knowing process. I do indeed believe
that we start with postulates and presuppositions, but, I add, we
start also with perceived truths. Now, a perceived truth in no way
presupposes a theory of any sort or kind. It is logically ultimate,
it is a crucial test of our theories and of their presuppositions. Even
a theory of the possibility of knowing presupposes its crucial tests ;

66 METAPHYSICS AND EPISTEMOLOGY

whereas if the crucial tests themselves presuppose the theory, they
are not crucial tests. To be sure, there is one philosopher for
whom this argument is utterly inadequate, namely, the believer in
the monistic or organic theory of truth. He really denies that
there is any such relation as logical priority ; and I have to confess
that I know no way of refuting the theory of a man who rejects
formal logic and holds consistently to this rejection. Fortunately,
however, the monist is never a consistent monist, for he argues ; and
it is, I believe, possible to reduce his monism to the absurdity
pointed out by Mr. Russell in the essay to which I have referred.
Here the realistic reader may again protest: "The dogmatist,
no matter where he starts logically, will sooner or later come to
the problems of the theory of knowledge; and he must solve these
epistemological problems in such a manner that his solution will
be consistent with the solution of his logically prior problems.
Therefore in his solution of logically prior problems he has already
indicated implicitly a part at least of his future epistemological
theory. If so, he really starts with an epistemological theory."
Yes, but only in the sense in which I can maintain that the chem-
ist starts with a biological theory. His chemistry has no doubt
shut out certain conceivable biological hypotheses which presup-
pose that present-day chemistry is false. So, no doubt, logic, math-
ematics, physics, and biology shut out certain epistemological
theories.1 Yet notice this is so, not because they presuppose any
theory of knowing, but because our theory of knowing presupposes
them. It is utterly idle to work out a theory of knowing which
presupposes that the exact sciences are false, unless we are pre-
pared to go back to intellectual savagedom. In short, we have to
start logically somewhere, but this 'somewhere' is not with a
theory of knowing ; and in so starting, we do indeed shut out all
theories of knowing which contradict whatever constitutes the
ultimate crucial tests of our theories. My own conviction is per-
ception is that ultimate crucial test, and as such it does not presup-

1 Personally I believe that they shut out idealism.

THE LOGICAL POSITION OF EPISTEMOLOGY 67

pose its own possibility. It simply is; and the man who questions
it assumes it in order to do the questioning.

5. If all of this be so, what follows ? In case our beliefs are en-
tirely built upon assumptions, then our whole question becomes :
Are the assumptions of epistemology more nearly fundamental
than those of other sciences; in other words, what is the logical
position of epistemology among the sciences? Whereas, in case
there are perceived truths, then either epistemology has to show
that it has a monopoly of these perceived truths and that it is in
its assumptions logically independent of the sciences, or it has to
admit that other sciences have no need of its good offices. But
this again is merely to say, epistemology has to show that it is
logically prior to the other sciences. That is, our whole question
regarding the science of the possibility of knowledge reduces to
the question, What is the logical position of epistemology among
the sciences, and, in particular, what is the logical position of that
branch of it which deals with the conditions of knowledge ? As
I have already shown that epistemology presupposes logic, my
question may be restated as the two following problems : What is
the logical position of the theory of knowledge relatively to the sci-
ences other than logic ? and, in particular, What is the logical po-
sition of that branch of epistemology which investigates the
possibility of knowledge ?

6. The answer to the first question asserts : not only is the
theory of knowledge subsequent to logic, but it is subsequent
also to some of the special sciences, such as physics and biology.
The knowing process as a natural event is conditioned by many
factors, the mind's physical and social environment on the one
hand, and the needs, the structure, and the health of the bodily
organism on the other hand. No explanation of knowledge has
ever been given that ignored totally what we know regarding
these factors.

In opposition to this, no one will claim that an epistemology
can be deduced from the principles and formulae of formal logic
alone. Surely more data must be allowed the investigator. But

68 METAPHYSICS AND EPISTEMOLOGY

what data? Would it be sufficient that the epistemologist be
furnished in addition with the sciences themselves as systems
of propositions in which various doctrines are demonstrated?
If so, his task would be to show precisely what are the logical foun-
dations of the sciences. That is to say, he would state rigorously
the ultimate premises presupposed in them, he would define the
fundamental notions as far as they are definable and point out the
ones that he finds himself unable to define. It is true that many
an epistemologist, e.g. Kant, has attempted to solve some of these
problems, which fall strictly within the field above defined ; but
is this epistemology ? It is not, for as we have seen it would ex-
clude a study of the knowing process, of the factors entering into
knowledge, and of the growth of knowledge and of many other
problems usually accredited to the theory of knowledge.1

Well then, would he have sufficient data if he got all his other
information by mere introspection, if he became an expert intro-
spective psychologist of the knowing process ? As such an in-
vestigator he would watch our knowing in the making; and he
would describe for us the facts precisely as he directly observes
them, for he would not derive them from physiology, physics,
or any other science. In this case he could not offer us any ex-
planation of these facts, or give us any hint as to what part they
play in man's life, as to what factors influence the knowing
process from without, as to what goal human cognition is prob-
ably heading toward, as to what limitations are set to the field

1 Of course if the reader chooses to limit thus his own use of the word, he may
have a right to do so ; but as the names metaphysics and epistemology are em-
ployed in this essay, the problem of the purely logical foundations of the sciences
falls entirely within metaphysics and entirely without epistemology. There is
indeed danger of idle dispute regarding the proper use of these words ; for recent
books entitled theory of knowledge and the great classic writings of the past two
hundred and fifty years often contain matter that is both metaphysical and episte-
mological as I define these terms. This difficulty, however, can and ought to be
avoided, for the nomenclature lies quite beyond the purpose of my essay ; whereas
the specific problems and the logical relations of their solutions are all-important.
Therefore the conclusion follows, as the word epistemology is here defined, that the
investigator in that field must have other or additional data.

THE LOGICAL POSITION OF EPISTEMOLOGY 69

of possible knowledge. Indeed, some philosophers would urge
that he could not tell us about anybody's knowing process ex-
cept his own; and other philosophers would maintain that he
could do even this only through outside information regarding
his own life and its environment. Evidently, the science in which
we are interested should not be thus hedged in; for, as we see,
the chief problems regarding the knowing process, its nature,
its conditions, its growth, its goal, and its limitations would
have to be excluded.

If, then, we are to require all this knowledge from the epistemolo-
gist, what sources of information must be open to him ? Clearly,
almost everything that modern knowledge can put at his dis-
posal. He must know all that biology and, in particular, physi-
ology can tell him of the relation between man's body, its func-
tions, and its origin on the one hand, and man's knowledge on the
other hand. He must know of the functional relationship be-
tween man's environment, both physical and social, and man's
knowledge, between our instinctive impulses, needs, and purposes
and our knowledge. He must know the history of our knowl-
edge and of our sciences from prehistoric days to our own tune,
in order to learn the course of scientific evolution and the fac-
tors which determine this course. Likewise, he must know the
development of knowledge in the individual and the factors which
determine it. In short, besides contributions from biology
and physiology, he will need all the help psychology, social psychol-
ogy, physics, political and social history, and the history of sci-
ence can give him.

Thus, one must answer our question regarding the logical posi-
tion of epistemology among the sciences somewhat as follows :
Epistemology is not logically fundamental ; on the contrary, it
presupposes logically the results of many of the special sciences.
If these results are false, so also, as far as we can tell, is epistemol-
ogy; and without these results granted as data the epistemolo-
gist would be unable to solve most, if not all, the problems belong-
ing properly within the field of his science.

70 METAPHYSICS AND EPISTEMOLOGY

7. Much the same is true of our second problem, the problem
we face when within epistemology we try to ascertain what knowl-
edge is possible and how it is possible. Man has never succeeded
in getting trustworthy information on this subject except em-
pirically ; for in case after case man has been able to discover what
scholars in an earlier age pronounced unknowable, or would
have pronounced unknowable if the question had so much as en-
tered their minds. This has been true of what man has learned,
precisely as a similar truth holds regarding what he has proved
himself able to do in spite of an earlier belief that the deed was
impossible. We have been able to see the far-distant and the ex-
ceeding small where centuries ago such vision would have seemed
impossible. We have been able to study the chemistry and the
temperature of the stars, we can weigh the planets, we can tell
with complete accuracy the area of curved figures whose sides
stretch out to infinity. In short, precisely as our wireless tele-
graph and telephone, our X-ray photographs, and our trolley
cars would seem miracles to Galileo, could we suddenly usher
him from the seventeenth century into the twentieth; so, too,
what has proved possible for man to learn since his day would seem
to him miraculous.

It will be objected that all such trite instances and the whole ar-
gument which appeals to them are entirely beside the issue. " What
within certain large areas will prove possible or impossible can of
course be learned only inductively; and Galileo might well have
been clever enough to refuse to answer questions regarding such
matters. But Galileo was able, and as able as we, to study di-
rectly the nature and conditions of knowledge and learn the ul-
timate boundary within which knowledge must keep. Thus,
Galileo could not predict the future of physical and astronomi-
cal science nor in any way give the details of what was to prove
possible within this general field which as a division of science
he knew to be possible; but he could have shown that there are
other problems essentially unlike any man has ever solved, and that
man lacks totally the kind of mind which could solve these prob-

THE LOGICAL POSITION OF EPISTEMOLOGY 71

lems. Take as an illustration a doctrine of Kant. Man has a
sensuous intuition only. He lacks an intellectual intuition ; and
since there are problems which could be solved only by such a
higher faculty, man must be content to let these problems remain
forever unsolved. Thus man cannot experience God, just be-
cause God is not sensuous. Again, man cannot trace back the world
to an origin or know it in its totality, because from the nature of
the mind such is not a possible experience. The two types of
problem are fundamentally distinct, and the methods of their
solution also are fundamentally different." Thus the criticist
agrees with the dogmatist that in the one case only an inductive
study of the history and status of the problems before the sci-
entist could enable us to predict with any probability what will
or will not prove possible knowledge : whereas, in the other case,
he maintains that a direct study of the nature of knowledge can
show that some problems are and some are not solvable. We must
accordingly turn our attention to the latter case to ascertain pre-
cisely what constitutes this branch of the theory of knowledge.

What is this study of knowledge which can reveal the field of
possible knowledge and its bounds? If the question means by
a study of knowledge a study of the knowing process and of the
factors conditioning it, then evidently we are referred again to one
of the special sciences, to the empirical study of the origin and
growth and function of knowledge in the individual and in the
race. This study is not a fundamental or peculiar science. Let me
take Kant's doctrine, not as proof of this, but as an illustration of
what I mean. In the first place, his ' Critique ' assumes that we
possess certain types of information, and it endeavors to show that
the mind must possess certain faculties in order to make this knowl-
edge possible. Kant's actual argument, to be sure, confuses this
with some other problems ; but in the main his reasoning shows
clearly that he is presupposing a definite psychological theory
and that his problem introduces only what lends itself to induc-
tive and empirical research. In short, it is psychological. In the
second place, Kant's 'Critique,' holding to a certain theory

72 METAPHYSICS AND EPISTEMOLOGY

of existence and to a certain theory of the knowing process, de-
duces from them the impossibility of our knowing a supersen-
suous world or the sensuous world in its totality. In short,
his results presuppose two things which make his theory far
from logically ultimate, a theory of existence and a psychology
of cognition. It is true that there is far more than this involved
in Kant's ' Critique ' ; but the moment we consider these further
problems, we pass over to a radically different sort of study of
knowledge and its possibility, for this study is not a study of the
knowing process proper, but of that which is known, the evidence
upon which it is based, and the postulates or axioms which it
assumes. Thus to return to our question, if we examine actual
specimens of the science of the possibility of knowledge, the
science that the criticist assures us is ultimate and sui generis,
we find only two things : first, psychology and cognate branches of
science, and secondly, a study of the logical foundations of our
knowledge. Either we find that the criticist is analyzing the
sciences logically to ascertain what data, or facts, what postulates
or principles, and what logical formulae make them possible;
or we are introduced anew to the historical and empirical study
of the origin and growth of knowledge in the individual and in the
race. If there is some further problem or some further method,
examination of the criticist's work fails to reveal it ; and the dog-
matist, in despair lest he has overlooked it, can only beg that it
be produced.

I am well aware that the criticist replies to all such statements:
"You do not understand criticism. It differs from other episte-
mologies precisely in its keeping psychology and epistemology
distinct. The mind, and the knowing which it studies, are not the
personal but the over-personal. Of course psychology is one of
the special sciences. This we too, not only admit, but teach, for
psychology itself presupposes epistemology." Yes indeed, it
is perfectly clear what your intentions are, and that if epistemol-
ogy would only be what you want it to be, it would indeed be
fundamental to psychology and in no way itself psychological.

THE LOGICAL POSITION OF EPISTEMOLOGY 73

But the question is not one of definition or of good intention.
The question is, What is the epistemology you offer mankind?
I know what you want to do, but do you do it ? Produce the episte-
mology that does not presuppose psychology. Produce the episte-
mology that is fundamental. My point is, the deed has never been
done and cannot be done; and therefore that there is but one
refutation to dogmatism, to wit, the production of an epistemol-
ogy that is at once truly fundamental and not a vicious circle. Most
dogmatists of to-day were brought up in the Kantian philosophy.
They had explained to them the nature of criticism, its funda-
mental character, and its difference from psychology. What is
more, they believed what they were taught. But later they began
to examine more critically the epistemology actually offered them
and found it not fundamental but saturated with scientific prej-
udices of one sort or another, found it distinctly a doctrine of the
day and generation of its author, found it, in short, a vicious circle.
To such a student of philosophy it is not enough to reiterate,
epistemology is such and such. There is but one thing to do,
that is, to produce the epistemology which is in accord with the
criticist's definition.

8. If the dogmatist is right in his conviction that the summary
aforegiven states fully the criticist's problem of the possibility
of knowledge, then the following seems to him a sufficient refu-
tation. As to the data, or facts open to possible observation,
history shows that human prediction is quite fallible, especially
where new methods and new instruments have come to man's aid.
As to our sensory and intellectual limitations, only the elaborate
empirical and inductive studies of the psychologist can give us
precise information; and psychology is one of the special sci-
ences posterior to several others. As to the postulates and
principles of science, history shows that these have often changed,
and experience has proved that not only in chemistry and phys-
ics, but even in mathematics, the method of trial and error has under-
lain man's discovery and selection of these fundamental proposi-
tions. Moreover, prediction here too has been decidedly fallible,

74

and no evidence whatsoever is forthcoming that any a priori
method of discovery will ascertain what principles will be suffi-
cient, not to mention which ones will prove necessary or even that
any will prove necessary. The history of philosophical mathe-
matics from Kant's time until to-day ought to banish from any
philosopher's mind the belief that the theory of knowledge can
reveal with certainty the necessary principles for the demonstra-
tions of the future scientist. Even the logician should learn from
past experience that many an accepted logical principle may in the
future prove to require radical revision, or at least may be capable
of further analysis or better formulation. Thus the science of
the possibility of knowledge is not sui generis, but is empirical and
inductive, as are most other sciences. The information that it
offers carries with it no categorical imperative to the special
scientist to be guided thereby ; for he is precisely the one that in
the past has successfully rebelled against the older or traditional
principles, and that has been the discoverer of the new ones which
take their place. Thus our general conclusion, I believe, must
stand : The theory of knowledge is not logically fundamental to the
sciences, and it cannot by any direct or a priori study of the know-
ing process ascertain the possible field or the limits of the sciences.

IV

EPISTEMOLOGY DOES NOT GIVE, BUT PRESUPPOSES, A THEORY OF

REALITY

1. WE have next to inquire whether or not the theory of
knowledge can give us general existential truths revealing the
outlines of reality or constituting a theory of reality. By main-
taining that the fundamental postulates, or principles of science, are
laws of thought, and that these laws of thought can be discovered
by the student of knowledge, transcendentalism claims to be able
to show a priori the most general features of the existent world.
Thus according to the familiar doctrine of Kant, the world we

A THEORY OF REALITY PRESUPPOSED 75

experience is determined in part by the experiencing mind, for the
mind's nature gives to whatever we experience its form. In this
way he explains that the world which we experience is a world in
space and time and a world ruled by causal law. Or, as we might
put it to-day: were our minds of a different nature, the world
which we should then perceive and know might be quite other than
our present world; for example, it might be a spaceless world,
or if a spatial world, it might have four dimensions ; or again,
if it were spatial and had three dimensions, it might none the less
be a world in which parallel lines meet or in which the sum of the
angles of a triangle are more or less than two right angles. In
short, we owe it to the nature of our mind that in this world the
sum of the angles of a plane triangle equals two right angles.
Hence, if our modern mathematicians can deduce for us various
geometries other than the familiar one of Euclid, the transcen-
dentalist can show us from the nature of our experience which one
of them all is the one that truly holds in the world about us. The
others may be perfectly logical, that is, true if their premises are
true; but the epistemologist shows that their premises are not
true. Moreover, to do this he does not have to go to nature, nor
does he have to experiment, measure, and observe until he finds
some facts inconsistent with all but one of the geometries ; rather,
he can employ the far easier method already described.1

Besides the Kantian and more closely related doctrines I wish
to include under the term transcendentalism the neo-Hegelian
theory of knowledge. Thus we are told by some Hegelians :
Reality is experience ; and reality is at least to some degree man's
experience, for though our experience falls short of completeness and
perfection, and though on this account we must look beyond man's
mind for the perfect experience of reality, still man's experience
reveals the essential nature of that perfect experience. Our ex-
perience develops, and the course of its development reveals the
goal whither it tends. That goal is the experience of the abso-

1 Cf. as a recent example of Kantian transcendentalism, Bauch, Bruno, Studien
zur Philosophic der exakten Wissenschaften, 108-141.

76 METAPHYSICS AND EPISTEMOLOGY

lute or universal mind, and this experience is absolute reality.
Briefly expressed, it is the business of the epistemologist to ascer-
tain from our knowledge the general nature of a perfect knowledge,
and from that to infer the general nature of absolute reality.

Thus from the nature of our knowledge it can be shown that no
self-contradictory experience is true, and from this it can be in-
ferred infallibly that reality is self-consistent. This example is
perhaps not so startling to the uninitiated as some others, because
it happens that no sane man believes that two contradictory
propositions can both be true. Yet for our purpose its seeming
self-evidence makes it an excellent illustration, for if the dog-
matist is in the right, even this argument of transcendentalism
is fallacious. Again, that reality is an organic unity, and that this
can be deduced from the nature of knowledge is another favorite
doctrine of some Hegelians. They argue : As our knowledge grows
at one part all other parts undergo change ; for even the proposi-
tion, two plus two equals four, is not the same truth to you and to
me now that it was in our childhood. As our insight into other
things mathematical has increased, so too has our insight into this
information belonging to childhood; and so the adult's state-
ment, two plus two equals four, is, strictly speaking, not the same
as the child's, for it has become a profounder knowledge. Hence if
knowledge grows as a totality and never by mere addition of new
information to the old, if its various parts are so organically con-
nected that a change anywhere means a change everywhere, and
if an increasing and better knowledge always reveals these aspects
even more prominently; then a completely true or perfect knowl-
edge must be a perfect organic unity. Hence each so-called part
will be what it is because of the whole, and the whole will be what
it is because of each part. Now if such a knowledge is the truth,
if, in other words, to be true it is compelled to be such a knowledge;
then the world also of which it is a perfect knowledge must be an
organic unity. The universe must be what it is because of its
members, and the members must be what they are because of the
whole to which they belong ; it cannot be a mere aggregate of in-

A THEORY OF REALITY PRESUPPOSED 77

dependent parts. Hence, again, if you ask the criticist how he
knows this, he does not reply that he has got his information in
experimental laboratories, astronomical observatories, or through
research in the field ; rather, he asserts it follows from the nature
of knowledge.

2. How persuasive is the argument, yet is it not utterly falla-
cious ? And if it is fallacious, is it not perhaps the most gigantic
case of self-deception of which the human intellect has been guilty ?
What does the dogmatist teach in opposition? One thing just
as dogmatist he does not do, and that is to deny any of the existen-
tial propositions of transcendentalism. It may be that the world
is a causal system, or it may be that the world is an organic unity.
What he does object to is the means by which the criticist claims to
get this information. He denies that the study of the nature of
knowledge can reveal any such theory of reality. Hence he
believes that if the criticist's theory of reality be true, it has been
logically smuggled into the theory of knowledge and then exhibited
afterward as a home product.

In particular the dogmatist's objections will differ somewhat for
different types of transcendentalism. Thus the transcendentalism
of Kant, and of those who follow him or Hume or Berkeley closely,
often asserts less general existential propositions than does the
present-day Hegelian transcendentalism. Now these less general
existential assertions are easy either to disprove or to trace back
to their logical origin in the special sciences; whereas the criti-
cism of such a proposition as the one asserting the world to be an
organic unity is more difficult both to formulate and to follow.

Let us consider first the objections to the former type of tran-
scendentalism. The attempts to discover a priori the structure
of reality within the field of the natural sciences has been, as al-
most all philosophers admit, most unsuccessful. Who to-day
would dare infer from the nature of knowledge the number of the
planets in our solar system ? Yet is it any the less foolhardy for
the philosopher, basing his opinions solely upon his theory of knowl-
edge, to deduce the nature of space and tune, to decide whether

78 METAPHYSICS AND EPISTEMOLOGY

mechanics or energetics is in the right, to show that we ought
to banish such notions as matter, empty space, and infinity from
our existential sciences, to deduce the persistence of force, or to
claim that events in nature are not related by causal law ? More-
over, where the attempt has been made and where the inference
seems to be quite correct, there remains still the question : Was
not the method actually used quite other than a priori f Did the
argument indeed keep quite within the theory of knowledge ? The
dogmatist believes it did not. Rather he believes the principles
were discovered by means of a logical analysis of the scientific
views of the day and were afterwards fallaciously shown to be
laws of thought or necessary forms of cognition. If this is the
course of procedure, transcendentalism follows logically the re-
sults of the special sciences, and it certainly is not in a position to
dictate to science her principles. Indeed, the position of transcen-
dentalism relatively to the sciences reminds one of the stern father
who ordered his small boy to go to bed, and upon the latter's reply,
"I won't !" said, "Then don't! I'll be obeyed." A growing science
is a hard youngster to discipline, and history shows that the
philosophers who have been foolhardy enough to lay down rules
for its behavior for all time to come have had later to beat a
retreat.

3. To be sure, merely to assert all this is not to prove it; yet
to prove it fully would require me to examine with great care many
different examples of critical epistemology. This cannot be done
here; rather it must suffice to take the greatest of the criticists
as our example and ask my questions regarding his work. Kant's
' Critique of Pure Reason ' claims to show that certain great exis-
tential principles are deducible from the nature of knowledge,
for they^ are i forms either of our intuition or of our understanding.
In other words, a study of the transcendental activity of the mind
will reveal the necessary principles of all existential science.

Whenc&did Kant get, and by 'get' I mean infer, these great laws
of the pure reason ? Did he, after genuinely studying the nature
of knowledge, derive his information from this study; did he not

A THEORY OF REALITY PRESUPPOSED 79

rather get his great principles first from his own scientific and meta-
physical research and from the science and metaphysics of his day,
and then did he not, after reading them into the nature of knowl-
edge, read them out again ? My conviction, and the conviction
surely of many students of Kant, is that the latter was altogether
the case; for is it possible to understand thoroughly the conclu-
sions of Kant's ' Critique of Pure Reason ' and be quite ignorant of
his scientific and metaphysical environment and precritical growth ?
Where did he get his phenomenalism, which is a premise and not a
conclusion of his argument? Where did he get his psychology,
which again is a premise and not a conclusion ? Indeed, is not his
psychology decidedly faulty, and has it not led him into many
epistemological and metaphysical errors? Where did he get his
first two antinomies which played such an important part in the
development of his transcendentalism? Did they come from a
study of knowledge or from a study of science? Moreover, is
there not good reason to believe that the Newtonian conception of
nature underlies many of his conclusions and often leads him to
think of the world of things in themselves as a follower of Locke
and Newton would have done ? l Further, were his conclu-
sions regarding the fallacious character of the arguments for the
existence of God genuinely the outcome of his study of the nature of
knowledge? Were they not rather the result of a profound study
of the arguments themselves ? Again, is his doctrine of space and
time truly the outcome of a direct and cautious study of our
spatial and temporal intuitions ? Is it not the other way round ?
That is, is not his doctrine of space and time an hypothesis logi-
cally "dependent upon his metaphysical conviction that mathe-

1 All the statements made in this essay regarding Kant are meant chiefly as
illustrations. To those who interpret Kant differently they may seem to call for an
extended exposition and proof, which would take us far from our main theme.
In justification of most of my views regarding Kant's Critique I shall then have
to refer to the excellent study of his theory of knowledge by Prichard. Kant's
Theory of Knowledge. Oxford. 1909. Cf. also Erdrnann, Benno. La critique
Kantienne de la connaissance comme synthese du rationalisme et de rempirisme.
Revue de Melaphysique et de Morale. 1904.

80 METAPHYSICS AND EPISTEMOLOGY

matics can furnish us infallible information regarding the nature
of real space and real time ? Finally, how about his doctrine of
causation? Surely the two following premises underlie logically
this part of his epistemology : first, the facts under observation in
the natural sciences do not reveal a causal relation or necessary se-
quence ; l and secondly, all explanations of nature presuppose
logically that the sequence of events in nature is a necessary order.
In other words, was Kant truly in doubt about the existential
validity of physics until by investigating knowledge he proved
to himself that experience would be impossible unless the under-
standing by a transcendental activity makes nature a causal sys-
tem; or was not this rather the logical order of his thought?
Nature is a causal system, physics is existentially true ; what
possible theory of knowledge then will account for the fact
that although nature does not reveal this causal-nexus to our
senses still we know that it is there ? Kant became here and there
an empiricist ; but we must not forget that he grew up and never
ceased wholly to be an old-fashioned rationalist, which indicates
that he remained a dogmatist in spite of his efforts to be a genu-
ine criticist. In short, his transcendentalism as a whole presup-
poses his precritical psychology and metaphysics, even though it
be true that here and there hi his ' Critique ' are to be found bril-
liant studies of the knowing process. If this be so, the Kantian
transcendentalism is a vicious circle ; 2 and that it is so, is no ar-
gument against the greatness of Kant, for even a Kant could not
do the impossible.

4. However, the transcendentalist can make his position far
more secure by reducing the existential principles he claims to
infer from epistemology to a few high generalizations; for ex-

1 As philosophers of the enlightenment beginning with Locke and Leibniz and
ending with Hume and Kant were coming to see more and more clearly.

* How apparent and utterly naive the vicious circle is in the phenomenalism of
Karl Pearson ! His ' metaphysics of the 'telephone exchange ' is almost explicitly
the presupposition instead of the conclusion of his theory of knowledge. In short,
by assuming a goodly supply of information regarding the transcendent world he
can prove to us that we can know nothing at all about that world !

A THEORY OF REALITY PRESUPPOSED 81

ample, if he teaches only these two propositions : reality is a
self-consistent system, that is, two contradictory existential propo-
sitions cannot both be true; and reality is an organic unity.
Whence does he infer these propositions, whose truth of course is
here in no way under debate ?

It is impossible to believe that the former proposition comes from
a study of knowledge. In the first place, the epistemologist be-
lieved this truth from his childhood and it seems most unlikely
that he avoids assuming the proposition as a premise in his episte-
mological research. In the second place, what is our proof that
two contradictory propositions cannot both be true? There
seem to be only four tenable answers : it is a self-evident truth
or axiom, it is a generalization from particular propositions, it
is an indemonstrable or ultimate assumption of formal logic,
it is a deduction of formal logic from some more nearly ultimate
postulates. In any .case, it is a proposition presupposed by a
large part of logic and is logically prior to any epistemological
investigation. It is not a law of thought. If in our thinking we use
it, we use it as a premise, and we use it because it is true. If our
thought is almost always compelled to use it, this is solely because it is
true, and because so few inferences fail to presuppose it. In short,
we can contradict ourselves, but we cannot contradict ourselves
and be correct ; and this is so, not because of the nature of thought,
but because two contradictory propositions are not both true.
Thus we may conclude : No examination of thought discovers the
law of contradiction or proves it, rather such an investigation
presupposes it. Hence if the world is a consistent system, our
thought has no more to do with making it so than has the nest-
building instinct of the oriole.

5. Again, is the world an organic unity? Let me grant it for
the sake of the present argument. Does this proposition follow
from the nature of thought ? The dogmatist denies that it does.
Rather the doctrine has its logical source in a principle usually
named in these days, the internal or organic theory of relations.
But whence this principle? Sometimes the philosopher who

82 METAPHYSICS AND EPISTEMOLOGY

holds it appears to infer it from the general results of the sciences.
If so, it is a generalization such as a law of physics. Some-
times the philosopher appears to infer it from a genuinely em-
pirical and inductive study of the evolution of knowledge hi the
race and in the individual. If so, again it is a generalization from
science.1 Now neither of these methods of discovery is consistent
with transcendentalism, nor can either be made to be so. There
seem to be but two ways of escape from this conclusion. The
first is, to retrace our steps and to show that the theory of knowl-
edge is fundamental and that reality is to be identified with the
knowing mind. The second is to admit frankly that this principle
is an ultimate postulate or axiom of a science logically prior to
epistemology; namely, a proposition of metaphysics. With the
latter dogmatism has no quarrel, for it frankly gives up transcen-
dentalism. With the former, however, the case is different. Its
first proposition is false. Its second, that reality is to be identi-
fied with the knowing mind, is certainly a proposition which can-
not be proved by epistemology, for epistemology presupposes too
many existential propositions from science to avoid a vicious circle
in any such proof. If, then, the epistemologist assumes it, he does
so as a metaphysician and a dogmatist.2

6. What then should be our general conclusion regarding tran-
scendentalism ? Transcendentalism stands or falls depending
upon the truth or falsity of two propositions: first, that highly
general information regarding reality can be inferred from the na-
ture of knowledge; secondly, that this information is truly fun-
damental, that it does not itself presuppose an array of exis-
tential generalizations and postulates borrowed from the other

1 Joachim in his book, The Nature of Truth, seems to me to do both.

1 Kant's boasted discovery, which he compares with that of Copernicus, comes
here in question. Is it an assumption or a valid conclusion of his Critique? If
the former, he is an out-and-out dogmatist. If the latter, he has wrought a logical
miracle ; for his epistemology certainly assumes existential propositions, and how
could it do so without assuming an existential proposition of such high generality
or the contradictory of this proposition ? It is gratuitous to add that I believe
he unconsciously presupposed the last.

THE VERDICT OF HISTORY 83

sciences. The evidence shows that the latter at least of these
two propositions is false. In short, transcendentalism is a vi-
cious circle. Its supporter pretends to derive from a stated source
information which unconsciously he has imported from elsewhere.
It is like a 'salted' mine, in which the most valued ore has been
put not by nature but by human hands. The result has been in-
evitable. Every intellectual enterprise except transcendentalism
seems to be prospering. We are learning much from the sciences
to-day regarding subjects that were once the center of the philoso-
pher's interest, the nature of the heavenly world, the nature of
matter, the nature of life, and the nature of mind. Even the
empirical and inductive study of the nature and growth of knowledge
is prospering, for we surely know far more about it than did Kant.
In contrast, what careful philosopher would offer mankind to-
day the amount of a priori information Kant claimed to derive
by means of his transcendentalism? Yes, the intervening years
have certainly proved the need of greater caution, and this too
in spite of an increasing insight into the nature of knowledge !
Is it not, then, high time for the epistemologist as well as the meta-
physician to declare his freedom from this difficult and fruitless
enterprise? There is no reason why a direct study of man's
growing knowledge and of the knowing process, if it be conducted
as one of the special sciences, should not yield great and valuable
and demonstrable results : but as long as the individual episte-
mologist feels it his duty to tease out of knowledge by a dialectic
a world-hypothesis, rather than to devote himself to a modest, open-
minded, and inductive study of cognitive facts, so long will his
work continue to promote intellectual distrust and to give back
disappointing rewards.

V

AN APPEAL TO THE PRAGMATIC TEST, TO THE VERDICT OP

HISTORY

1. THERE remains one further line of argument which must
not be totally neglected, the appeal to the pragmatic test, to the

84 METAPHYSICS AND EPISTEMOLOGY

verdict of history. Has epistemology been psychologically and
historically the chief source of metaphysical problems and of their
solution? During the past two centuries, in which the influence
of epistemology has been so great and in which she has had as
her leaders the ablest philosophical thinkers, has the progress of
metaphysics been due chiefly to epistemology rather than to the
special sciences ? Have two centuries of its dominance in philo-
sophical research been a help or a hindrance ?

It would be most unconvincing to offer a brief answer to these
questions, if I pretended for a moment that my answer were based
upon an analytic and well-established solution of the minuter
problems belonging to the history of philosophy. The same phi-
losopher has usually been both metaphysician and epistemolo-
gist; and the actual course of his thought from day to day has
seldom kept the two sets of problems distinct, but has interwoven
them most intricately. Then, too, his writings may completely
conceal the actual evolution of his thought. However, history
is written hi large letters as well as in small; and it is therefore
not impossible to make a brief and convincing statement regard-
ing the influence of epistemology upon metaphysics in the last
two hundred years.

What has done most to change our modern theory of reality?
To what discoveries or doctrines of the past two hundred years
is our present-day metaphysics especially indebted, to episte-
mology or to the progress of the natural sciences? Most de-
cidedly the latter. Even epistemology itself is similarly indebted.
How great a change in our conception of the world has taken place !
Not as great to be sure as the change from the thought of Dante
to that of Sir Isaac Newton; still the two changes are comparable.
Mathematics, physics, and biology have undergone a very great
and wonderful growth. Chemistry has been born and reached
maturity. So, too, has the historical research into almost every
field of human interest. That all this could have taken place
without influencing directly and profoundly our metaphysical
views is unbelievable. Let us see.

THE VERDICT OF HISTORY 85

2. First there has been the great growth of mathematical knowl-
edge. In the past one hundred years this has quite changed our
view of the nature of mathematics itself ; and, what is especially
important, it has done away with the older metaphysics of space.
Instead of a definite and infallible conception of the nature of
existent space and time, such as Kant believed we possess, we know
to-day that their nature cannot be inferred solely from pure
mathematics, but must be learned in part at least empirically
and inductively as truly as must the nature of light or electricity.
To be sure, mathematics also will continue to contribute very largely
to this knowledge, but mathematics contributes also very largely
to our knowledge of light and electricity. In short, mathe-
matics, and mathematics quite divorced from any epistemolog-
ical considerations, has completely transformed this old and im-
portant metaphysical problem, the nature of space and time.

A similar truth holds regarding the problem of Kant's first two
antinomies. The problem of the nature of the mathematical in-
finite and continuum certainly seems to have reached a new and
higher stage; and the resulting insight into these two notions,
together with a better understanding of the nature of mathe-
matics itself, have removed one chief source of error in the older
metaphysics. For, as pure mathematics alone cannot solve the
problem of space and time, so, too, it cannot solve the problem of
the origin, extent, or continuity of the physical world. This prob-
lem, as so many others, must remain unsolved until facts are
discovered which can be shown to form a crucial test of the merits
of rival hypotheses. Whereas if mathematics could furnish us
knowledge of the existent world without the aid of other sciences,
as Kant believed it could, then indeed a purely dialectical pro-
cedure might give us the information sought in this way by the
older dogmatic metaphysics. Indeed, one of the greatest philo-
sophical discoveries of all time seems to have been made, and
made in the nineteenth century; namely, the discovery that mathe-
matics is a non-existential science ; and this discovery we owe not
to the epistemologist but to the philosophical mathematician.

86 METAPHYSICS AND EPISTEMOLOGY

3. Not only are we indebted to mathematics, but also to physics
and chemistry for vast changes in our conception of the physical
world. Of all metaphysical problems the nature of matter is one
of the oldest. Now in the last few years, as we all know, we seem
to be learning more concerning the nature of matter than man
succeeded in discovering in the preceding two thousand years.
Even such a good old conviction as that mass is an absolute con-
stant is now contradicted, and what could be more startling than
to be told that electricity is to be an all but fundamental concept
in the new philosophy of nature? A thousand years of tran-
scendentalism or of any other theory as to what matter must be
in order to be a possible experience, could not have revealed to us
such truths. Again, the rise of thermodynamics and its doctrines
regarding the conservation of energy and the irreversibility of
nature's processes have modified greatly our conception of the
physical world about us. It may be that their ultimate meta-
physical significance is still hidden from us, but in any case their
great importance to metaphysics seems to be assured.

4. In the third place, nothing during the past one hundred years
has transformed more remarkably our theory of nature and of
life than has the doctrine of animal and plant evolution and in
general the modern historical point of view. To whom do we
owe this new insight ? In part, indeed, to men whose names are
foremost in the list of epistemologists, to Kant and Hegel, and to
men whom they have strongly influenced. However, it is very
easy to exaggerate this truth by inferring that we owe this part of
their contribution to their epistemology. Moreover, without any
desire to minimize our debt to them, there is every reason to be-
lieve that the evolutionary and historical point of view would
have come in the nineteenth century had they never lived and
had epistemology been completely neglected in those days; for
the men and the influences that led us to the new way of think-
ing belong to almost every department of European science and
go back probably to the days of Galileo. There is the growth of
astronomical theory from the Renascence to Kant and Laplace

THE VERDICT OF HISTORY 87

with their evolutionary hypotheses of the origin of our solar system.
There is the new and profound interest in history and historical
research due on the one hand to linguistic and literary discoveries
and on the other to the political, social, and religious ferment of
the times and, we should add, due in general to the new and wide-
spread romantic interest in the life of past ages and of other lands.
Finally comes the rise of the biological evolutionary hypothesis
with its tremendous influence upon the thought of our day. Surely
this doctrine also was chiefly due, not to the philosophical in-
quiries of the eighteenth century, but to the enormous accumu-
lation of geological and biological data which was compelling the
scientist to seek a theory to explain and to systematize them.
Indeed, one must infer in all departments that instead of our new
historical point of view being indebted to epistemology, episte-
mology has itself been completely transformed by this influence
from without. The most superficial study of the epistemology
taught to-day, be it that of the modern Hegelian or that of James
and Dewey, will reveal the truth of this conclusion; for knowl-
edge is now regarded by all as an essentially evolutionary process.
5. As against these changes in our conception of nature and of
life there stand five important metaphysical doctrines which
appear to be indebted especially to the study of knowledge:
first, the doctrine of primary and secondasy qualities; secondly
and thirdly, the eighteenth century's criticism of the older doc-
trine of causation and of substance; fourthly, the idealistic, or
spiritualistic, theory of reality; and fifthly, in more recent days,
the issue regarding the organic, or internal, theory of relations
leading to the opposed theories, monism and pluralism. Some
of these doctrines and their logical relation to epistemology have
been already considered; but we are now concerned solely with
influences, psychological and historical. The dogmatist has to
admit that the influence of epistemology here has been very great.
However, he makes this admission with a satanic delight, for he
believes that the influence has been all but disastrous. The in-
tellectual world has been led astray for over a century by a com-

88 METAPHYSICS AND EPISTEMOLOGY

•
plete confusion of two fundamentally different problems and by

the resulting mad hope that under the leadership of psychology
metaphysics was to find its way at last into the promised land.
Even our greatest philosophical thinkers for two centuries have
been under the spell. None the less much can be said to show that
even within these five problems other influences than epistemology
have been at work.

In the theories of the primary and secondary qualities there is
clear evidence that the physical doctrines of the day had their
strong influence upon all who held the theories. Moreover, there is
ample evidence to show that physics really went on its own way
minding its own business and neglecting what the epistemologists
had to say on this subject. Indeed, what physicist to-day need or
would bother his head with epistemological doctrines respecting
what are and what are not primary qualities ! This he learns by
questioning nature and by ascertaining what theories of matter
will account for the facts he observes. Indeed, the metaphysician
who to-day goes for information on this subject either to the psy-
chologist or to the epistemologist is liable to find that not one
in ten can tell him what physics has to say regarding the primary
qualities of matter, let alone, give him a theory at all adequate
in the light of our present knowledge and therefore worth
listening to. Rather what he will get as an answer is a vestige
pointing back to the physics of an earlier century. Perhaps no-
where has a combination of antiquated physics and epistemo-
logical metaphysics given rise to more worthless discussion. The
truth of the case is, we do not know what are the primary quali-
ties of matter. Energetics gives us a very different answer from
that given by the mechanist. Then, too, we are in the midst of a
period when a flood of new light regarding the ultimate nature of
matter is coming into physics.1 Indeed, the whole subject needs
to be studied anew from top to bottom by the metaphysician un-

1 Cf . on one physicist's view as to what is the basis for dividing qualities into
primary and secondary, Duhem, P., La tMorie physique. Paris, 1906. Pt. 2,
Chap. 2.

THE VERDICT OF HISTORY 89

der the instruction of the physicist. But all of this is beside our
question except in so far as it shows that an older physics has been
strongly influencing the views of many an epistemologist from
Locke's day to our own.

6. As regards the various doctrines of causation the influences
at work have probably been much more complicated than the text-
books on the history of philosophy indicate. It is true that as
we read Berkeley, Hume, Kant, and Mill, we are liable to feel
that epistemological considerations alone are at work behind
their criticism of the older doctrines of causation. The same feel-
ing, too, may come in reading Mach and Pearson. Still it is far
from certain that this feeling is justified, for they were dealing
with a problem that, though usually regarded as epistemological,
may in truth turn out not to be such. To be more explicit and
to illustrate : If I am looking at a landscape gorgeous with the light
of the setting sun in order to ascertain whether or not certain colors
are there, is my problem epistemological ? No, it is not, though it
be true that the step from my problem to those of epistemology
might be a very short one for many thinkers. Now the problem
of Berkeley and Hume and many other classical writers since their
day was in part one that is strictly analogous to the foregoing.
They were not studying the knowing process so much as the
actual empirical evidence of a necessary sequence of events. In-
spection, Hume really tells us, does not reveal any such connec-
tion. It reveals a succession, but not a causal relation. So the
nominalist Pearson of our own day inspects nature and does not
see there any causal law. He does see the flow of events, but he
claims that is all nature reveals. Hence, he concludes, to assert
the existence of law in nature is to read into nature a quite foreign
set of relations. This leaves him, and Hume before him, with the
problem, what, then, is so-called natural law and why do we tend
to talk as though it were a genuine element in nature ? In short,
quite apart from any question as to what is the correct solution
of their problem, it does seem as though something in addition
to merely epistemological influences were at work and that this

90 METAPHYSICS AND EPISTEMOLOGY

something else can be described as a reaction against older meta-
physical theories, together with a greater open-mindedness in their
direct observation of the facts of nature. If this is true, then
we have to conclude that a clearer awareness of the absence of cer-
tain facts, a keener insight into the logical topheaviness of the
theories of the day, as well as a greater attention to the nature
of knowledge, were all at work in bringing about the reaction
against the seventeenth-century doctrine of causation.

7. In regard to the doctrine of substance, a much stronger proof
can be offered that direct evidence in the form of fact, together with
the progress of theory in science, played an important part in the
change from seventeenth-century thought, for the old notion was
rejected not only by philosophers, but also by scientists. In meta-
physics the substance hypothesis had led to views as far apart as
materialism and occasionalism, as Spinoza's monism and Leibniz'
monadism; and this divergence of opinion certainly promoted a
skeptical attitude toward the whole endeavor to explain the
world in terms of substance. In natural science the tendency was
more and more to lay stress upon the relations between things,
and less and less upon substance and attributes; in fact, modern
physics was largely a reaction against just this notion, the old no-
tion of forms. Especially does this change of view come out in
the downright hostility to any explanation which makes use of
the notion of force or forces.1 In short, the growing explicit oppo-
sition to the old doctrine of substance not only had to come,
epistemology or no epistemology, but did come through many
influences other than the study of knowledge.

Moreover, the Kantian and post-Kantian epistemology has
been rather a conservative influence against the tendencies of the
natural scientists and of some metaphysicians. Though Kant
admits the full justice of Hume's criticism of the dogmatic doc-
trine of substance, he endeavors to show that substance is a neces-
sary form of thought and has validity a priori; and the modern

1 Cf. the opposition to Newton's theory of gravitation on the part of the Car-
tesians.

THE VERDICT OF HISTORY 91

members of the Kantian-Hegelian schools, in their doctrine of the
absolute, tend strongly to uphold the substance-attribute notion
as the fundamental notion in the theory of reality. Hence,
whether right or wrong, the modern tendency within science to op-
pose any use of the substance notion and to confine the proposi-
tions of science to assertions of relations between terms, has
little indebtedness at least to German epistemology.

8. In regard to the two remaining metaphysical issues, the issue
between those who hold to the organic theory of reality, or pres-
ent-day monism, and their opponents, the pluralists, and again
the issue between the idealistic spiritualists and those who find
their views unwarranted, it must be frankly admitted that the
influence of epistemology has been very great indeed. But
here again the question transforms itself into the other question,
has this influence been for good or for bad ? — a question, however,
which lies beyond the field of our present inquiry. Here, then, the
dogmatist must be content to urge his conviction that such doc-
trines should be based by their upholders upon facts and princi-
ples that may indeed be presupposed by a particular epistemology,
but that themselves are truly fundamental. Until this is done
monism, pluralism, or any other metaphysical theory is a house
built upon the sands.

9. I have appealed to history, and the answer seems on the whole
to be decidedly in favor of the dogmatist's prejudices except in
the case of certain prominent metaphysical doctrines of to-day,
which are explicitly founded by their advocates upon episte-
mological considerations. But even here the dogmatist finds
the influence of many non-epistemological factors, such as the
doctrines of evolutionary biology and psychology. Hence the fol-
lowing conclusion seems to be just : Where the change in our
modern views of nature and of mind is admitted by all to have
been genuine progress, there epistemology has not played the
part it should if it be the truly fundamental science; for this prog-
ress has come notoriously from other sources and has been very
influential even in bringing about changes within epistemology

92 METAPHYSICS AND EPISTEMOLOGY

itself. On the other hand, where no doubt exists that episte-
mology has been supreme in its influence, there one finds to-
day the most serious questioning as to whether the influence has
been good or bad.

VI

METAPHYSICS SHOULD BE EMANCIPATED FROM EPISTEMOLOGY

1. WE are now prepared to take up the general question,
Should not metaphysics be emancipated from epistemology ? If
epistemology is not logically fundamental, if epistemology cannot
of itself show either what knowledge is possible or how knowledge
is possible, and finally if epistemology cannot give us the logical
foundations of a theory of reality, may we not conclude that
metaphysics owes neither its problems nor their solution es-
pecially to epistemology? I believe that we may, for I am con-
vinced that all the reasons for making metaphysics identical with
epistemology or logically and methodologically dependent upon it
are those which have been given and disputed.

Metaphysics as a logical study of the foundations of the sciences
needs as its data only two things, the sciences in their most rigor-
ous formulation and formal logic. What metaphysics as a theory
of reality needs may be more doubtful. The following may
serve as a brief and tentative answer. It needs the preceding
branch of metaphysics, for that study will reveal the theory of
reality implicitly contained in the sciences; and with this meta-
physics will certainly have to reckon. Will it need more than
this? That depends upon how the following difficult questions
are answered, questions which themselves perhaps belong quite
within metaphysics. In the first place, will not logical analysis
reveal besides the foundations of science the foundations of other
independent systems of propositions, systems at least implicitly
asserted in man's art, in his morals, and in his religion ? If so,
will not the theory of reality have to presuppose them? Let
me reply, Yes. Secondly, what is the factual, or that ultimate,

METAPHYSICS SHOULD BE EMANCIPATED 93

concrete which we observe but do not either assume or infer?
Can there be a science just of it, a 'Gegenstandstheorie,' if you
will? Is this ultimate truly analyzable; or is it alogical, as
Bergson and James believe ? Here let me reply : If there be such
a group of problems, either they are a part of metaphysics and not
of some more nearly ultimate science upon which metaphysics
depends ; or they and their solution fall within the bounds of the
different special sciences and of the other aforementioned non-
scientific systems. In the former case, our non-metaphysical
systems will presuppose our metaphysics, whereas in the latter
case metaphysics will arrive at this body of information by logi-
cal analysis of all these systems. Probably this is not a genuine
disjunction, and both propositions are in part true. However, all
of this is a matter not of theoretical but of great methodologi-
cal importance; for it reduces to the purely methodological
question, who is the real metaphysician, the real authority in
metaphysics? Is he mankind at large or is he the professional
metaphysician? Finally there is a third question: May not
metaphysics have among its postulates or indemonstrable propo-
sitions some that are nowhere else to be found, that are peculiar
to metaphysics? If so, these parts of metaphysics are certainly
fundamental. In short, all of these many problems (including
logic) are fundamental, and the sum of their solutions certainly
constitutes a science which underlies logically the remainder of
our knowledge. It is the first science; and if so, it is meta-
physics.

Further we may make the following general statements regard-
ing the methodological indebtedness of metaphysics to other
bodies of knowledge. In parts it may be indebted to none. In
great part it is surely indebted to logic. As a study of the logical
foundations of the sciences, it has to be given the sciences themselves
as data. Finally as the theory of reality, it may draw information
and help from every department of man's intellectual life. The
growth of science can revolutionize metaphysics as it did in the
days of Galileo. A radical change in one or more of the postu-

94 METAPHYSICS AND EPISTEMOLOGY

lates of science may do the same, and so also may a great empiri-
cal discovery. Then, too, as man grows in artistic taste, and in
moral and religious insight, he discovers new and, it may be,
fundamental truths. If so, his discovery may lead him to revise
radically his theory of reality or at least to present to the meta-
physician new problems for the latter to solve.

2. Finally we come to the special question, what does meta-
physics owe to epistemology ? It owes much by way of sugges-
tion. The story of how human knowledge grows in the individual
and in the race, and the story of the knowing process itself do
often suggest to the student of logical analysis what to look for
and where to find what he seeks. For example, that knowledge
grows by the trial and error method, or the experimental method,
indicates at once that science as a system of propositions has as its
premises many unproved assumptions, postulates, and guesses.
The story of this growth indicates also that the work of the meta-
physician will never be finished as long as man keeps growing in-
tellectually; for no sooner do we work out (as the philosophers
did in the Middle Ages) the theory of reality presupposed in or
consistent with the knowledge of one age than the work has to
be done once more for the knowledge of a new age. A second
example is the great mental law of association which suggests at
once (as it did to Hume and to Kant) that causation must be one
of the fundamental or nearly fundamental postulates of science;
and, to take a third example, the theory of knowledge can throw
much light upon the history of science, and the history of science
in turn is full of suggestions for the metaphysician.

3. All this is true of the methodological relationship of epis-
temology to metaphysics, and it would be foolish indeed not to
admit it ; but at the same time we must not forget the limitations
of this indebtedness, for the errors that can then arise are, as we
have seen, very serious. A correct epistemology can be full of
valuable suggestion to the metaphysician; but this science is in
no peculiar respect nor to any peculiar extent fundamental to
metaphysics. It is not peculiarly a part of metaphysics, nor is

METAPHYSICS SHOULD BE EMANCIPATED 95

it in any respect to be identified with metaphysics. On the con-
trary I have endeavored to show that epistemology is not a logi-
cally fundamental science, that the solution of the problem of
the possibility and the limits of knowledge is logically subsequent
to some at least of the special sciences, that epistemology cannot
furnish us with a theory of reality, that metaphysics owes logi-
cally neither its problems nor their solution to the theory of knowl-
edge, and finally that though the full verdict of history has not
been delivered, there is strong evidence that criticism has se-
riously hindered as well as helped metaphysics during the past
two centuries. If these conclusions are true, then metaphysics is
by right free and independent of epistemology and should at once
proceed to emancipate itself entirely from the dominion of this
science.1

1 Though few physicians are expert in diagnosing their own case, still I can
perhaps help some readers to discover my bias or prejudices. As we look back over
the course of this argument, what is the world of discourse within which it proceeds ?
The answer is, logical analysis. There are, I believe, two prominent and radically
different points of departure nowadays in our philosophical studies. One man is im-
pressed with the facts of psychology; and though he admits that psychology itself
is one of the special sciences, he still seeks a philosophical foundation by means of a
study of these facts. The other man, though not blind to these facts, cannot regard
them as the most significant ; rather he is impressed with the truth that the chief
business of science is to demonstrate. As a consequence, the question, What are
the premises of any hypothesis ? is the all-important philosophical problem. Every-
where in man's knowledge he finds two sets of premises, on the one hand the prin-
ciples of formal logic and on the other hand postulates and observed truths, or facts.
Now the sum total of these presuppositions form the philosophical foundation upon
which he believes he must build; and, of course, he sees in the other philosopher a
thinker who has hopelessly confused psychology and logic. In short, the one man
is temperamentally a psychologist; the other, a logician. Moreover, each feels that
he has dug deeper down for his foundation than has the other. Hence the dead-
lock where neither seems able to convince the other, and where our ultimate duty
to one another seems to be not to try to persuade or to refute but to try to make
our positions clear to each other. Even though each party may see truths to which
the other is blind, and even though the ultimate verdict may be a part victory for
both, I am convinced that the two positions have their centers at different points,
and that these points are mutually repellent. If so, one is essentially right and
the other IB essentially wrong, and there can be no compromise.

A REALISTIC THEORY OF INDEPENDENCE

A REALISTIC THEORY OF INDEPENDENCE

BY RALPH BARTON PERRY

I

THE IMPORTANCE OF THE NOTION OF INDEPENDENCE

EIGHT years ago, in describing "a curious unrest in the philo-
sophic atmosphere of the time," Professor James concluded with
the remark that "strangest of all, natural realism, so long decently
buried, raises its head above the turf, and finds glad hands out-
stretched from the most unlikely quarters to help it to its feet
again." l This reanimated corpse is now fairly on its feet, and able
to protest with Mark Twain that the reports of its death were
"greatly exaggerated." As a living and hopeful member of the
philosophical community, it is naturally concerned that its iden-
tity should be unmistakably defined, lest it should again be care-
lessly interred or reported missing; or lest in the me!6e of con-
troversy it should suffer from blows intended for another. The
present essay attempts such an identification of realism redivivus —
or of what may now conveniently be designated neo-realism.

1. It is necessary, in the first place, to explain the crucial im-
portance of the conception of 'independence.' For the term
'realism' is also traditionally associated with another conception,
the conception, namely, of 'substance.' When construed in this
latter sense, realism is contrasted with 'phenomenalism,' 'immedi-
atism ' and ' empiricism.' It is taken to mean that the real is not
what is experienced, but some substance or essence which lies behind
what is experienced. The real, according to this view, is not con-
stituted by its predicates, but is manifested in them; it is the

1 James, W. Essays in Radical Empiricism, 39-40.
99

100 A REALISTIC THEORY OF INDEPENDENCE

subject that owns them, the ground that supports them, or the
cause that produces them. According to 'phenomenalism,' 'im-
mediatism,' or 'empiricism,' reality coincides with appearance —
things are what they are "known as." l According to realism in
the contrasted sense, reality is that which appears, that of which
something is known.

2. The most straightforward statement of this version of
realism is to be found in the following passage from Thomas
Reid's "Essays on the Intellectual Powers of Man": "Things
which may exist by themselves, and do not necessarily suppose the
existence of anything else, are called substances; and, with rela-
tion to the qualities or attributes that belong to them, they are
called the subjects of such qualities or attributes. All the things
which we immediately perceive by our senses, and all the things we
are conscious of, are things which must be in something else, as their
subject. Thus by my senses, I perceive figure, color, hardness,
softness, motion, resistance, and such like things. But these are
qualities and must necessarily be in something that is figured,
colored, hard or soft, that moves, or resists. It is not to these
qualities, but to that which is the subject of them, that we give
the name of body. ... In like manner, the things I am conscious
of, such as thought, reasoning, desire, necessarily suppose something
that thinks, that reasons, that desires. We do not give the name of
mind to thought, reason, or desire ; but to that being which thinks,
which reasons, and which desires." 2

The peculiarity of such a realism as this lies in the absolute dis-
tinction between real body and real mind, as substances, and the
'qualities' or 'attributes' by which they are known. The 'reals'
are different from the content of knowledge. Whether they are
also independent is a second and more ambiguous question. That
body is independent of mind would seem to be clear, in that as

1 Cf. James, W., Pragmatism, 50.

1 Reid, T., Essays on the Intellectual Powers of Man, Essay I, Ch. II, Hamil-
ton's edition (1895), 232. For an account of the development of the same general
view in recent German philosophy, cf. Stein, L., PhUosophische Strdmungen der
Gegenwart, Ch. VI.

THE NOTION OF INDEPENDENCE 101

substances bodies "may exist by themselves," while bodily qualities
"must necessarily be in" bodily substances. The qualities them-
selves, whether bodily or mental, are evidently not independent
in that they "must be in something else." Whether the sub-
stances themselves are independent of the qualities that are in
them is more doubtful. Locke, whose view of substance closely
resembles Reid's, regards it as probable that the soul may exist
without thinking, that its nature, in other words, is independent of
those forms of consciousness by which it is known.1 That the same
view might be held with reference to bodies is suggested by Locke's
repeated assertion that "the real essences, on which depend their
properties and operations, are unknown to us." 2

It is probably safer, however, to conclude that for both Locke and
Reid, some properties, namely, the 'primary' qualities, extension,
hardness, etc., belong necessarily to the nature of the body. There
is a notable difference between these two authors as respects the
ground of the distinction between 'primary' and 'secondary'
qualities. For Locke, the primary qualities are such as produce
similar ideas in the mind. Thus the quality 'hardness,' and the
'idea of hardness' produced by it, are similar. The secondary
qualities are such as produce dissimilar ideas in the mind, as when
"a violet by the impulse of such insensible particles of matter of
peculiar figures and bulks, and in different degrees and modifica-
tions of their motions, causes the ideas of the blue color and sweet
scent of that flower to be produced in our minds." 3 Reid, on the
other hand, sought to avoid that "theory of ideas" which he held
to be the besetting sin of philosophy. Knowledge is not a having
of ideas produced in the mind by things. It is an act of belief in

1 Cf. Essay concerning Human Understanding, Book II, Ch. I, §§ 10 ff. For
Locke's view of substance, cf . Controversy with the Bishop of Worcester, No. IV
(St. John's Edition, Vol. II, 352) : "We cannot conceive how simple ideas of sen-
sible qualities should subsist alone, and, therefore, we suppose them to exist in,
and to be supported by, some common subject ; which support we denote by the
name substance."

2 Essay concerning Human Understanding, Book IV, Ch. "VT, § 12.
» Op. cit., Book II, Ch. VIII, § 13. Cf. §§ 7-26, passim.

OK

102 A REALISTIC THEORY OF INDEPENDENCE

an object, of which one must have a 'conception.' Precisely how
Reid distinguished 'conception' from 'idea' is far from clear.
But it is evident that he wished to get rid of the traditional barrier
between the mind and its objects. The 'conception' was an act or
instrument of the mind itself, and not a product or counterpart of the
things. It was part of that faculty of knowing which the author
was satisfied to leave inexplicable since he was confident that it
must in any case be presupposed.1 In any case primary and
secondary qualities are distinguished without reference to 'ideas,'
the former being the qualities that are known directly and dis-
tinctly, the qualities of which one knows "what they are in them-
selves." "Therefore," says Reid, "were I to make a division of
the qualities of bodies as they appear to our senses, I would divide
them first into those that are manifest and those that are occult."
Reid's ' manifest ' qualities correspond to Locke's primary qualities,
such as 'extension,' 'hardness,' etc.; and his 'occult' qualities
comprise Locke's secondary qualities, together with the feelings
which bodies induce hi the organism and the 'powers' which they
display in their operations on one another.2

Thus with Reid the qualities of bodies, whether manifest or oc-
cult, belong to the bodies themselves. They are not primarily
'ideas,' of which some are similar and some dissimilar to bodies.
Their original locus is in the bodies themselves. There can, there-
fore, be no question of their remaining mere ideas through the un-
certainty of the existence of their bodily counterparts. In other
words, Reid believes himself to have removed the assumption
which underlies the idealism of Berkeley and Hume. And yet the
realism of Reid is open to a very obvious idealistic rejoinder. For
the qualities do not constitute the body. He makes it perfectly
evident that the principle of substance is the controlling motive
in his thought. He thinks that " it requires some ripeness of under-

1 Essays on the Intellectual Powers of Man, Essay II, Ch. XX. For the
difficulties and ambiguities of Reid's view, cf. Sir William Hamilton's notes to his
edition of this and other essays.

* Op. cit. (1895), 313, 322.

THE NOTION OF INDEPENDENCE 103

standing to distinguish the qualities of a body from the body."
The relation of the qualities to the bodily "substratum" must
doubtless remain "obscure"; it "is not, however, so dark but
that it is easily distinguished from all other relations." 1 The
qualities of bodies are thus left in a precarious situation. Since
they are not identified with the body, they may the more easily
be captured by mind and converted into ideas; and since body
bereft of them is reduced to a nullity, it may the more easily be
ignored as a non-entity. Thus, the principle of substance betrays
realism into the hands of its enemy.

3. It is reasonably clear, then, that the traditional realism has
been both confused and compromised by an alliance with substan-
tialism. In view of this fact, the critics of realism are scarcely to
be blamed if they have not shown a nicety of discrimination which
realists themselves have failed to show. Of contemporary critics,
Professor Royce is especially notable for expressly identifying
realism with the theory of independence. Nevertheless, even this
writer has not succeeded wholly in separating this theory from
substantialism. Thus he distinguishes three "popular ontological
predicates." "To be immediate, or, on the other hand, to be well
founded in what is not immediate, and, thirdly, to be genuine and
true, — these seem to be the three principal conceptions of what it
is to be real in the popular ontology." The author presently con-
cludes that realism is "a synthesis of the three popular ontological
predicates, although, as history shows, with a preference for the
second predicate." In other words, "realism is fond of sub-
• stances, of 'inner' or of 'deeper' fundamental facts, and of inacces-
sible universes." 2 Now it happens that the realism of the present
day has strong aversion for these things. It is in sympathy with
the whole modern trend of thought toward identifying reality with
the elements, processes, and systems of experience. But it main-
tains that these elements, processes, and systems are independent

» Ibid., 323.

2 Royce, J., The World and the Individual, First Series, 54-55, 68 ; cf. also 63,
66, 67, 86, 106, 115.

104 A REALISTIC THEORY OF INDEPENDENCE

of being experienced. Although they may compose or enter into an
experience, they need not do so. In other words, neo-realism as-
serts the independence of the experienced on the act of experience ;
or of the sensible and intelligible properties of things on the opera-
tions of sensation and intellection. Thus realism must purify the
notion of independence of all suggestions of other-ness, remote-
ness, or inaccessibility, not only for the sake of a full and forcible
presentation of its case, but even to avoid being confused with a
whole alien and objectionable tendency of thought.

4. It must be confessed that realists have not as yet taken pains
to define 'independence.' Thus in a discussion before the Aristo-
telian Society of the question, "Are Secondary Qualities Independ-
ent of Perception," Mr. Nunn, as a realist, adopts the "affirmative
answer." 1 And in reply, Mr. Schiller justly puts the question,
"What does independent mean to a realist?" But as Mr.
Schiller himself remarks, the idealist has been no more precise in
his use of the term than the realist. Both Royce and Joachim,
who, like Royce, explicitly identifies realism with the independence
theory, constantly employ the term in their polemic without under-
taking to define it. Both of these writers characterize an independ-
ent entity as that to which another entity "makes no difference." 2
But this is only a figurative paraphrase of the term. It introduces
practical or dynamical considerations which are more confusing
than clarifying.

Nor can the idealist be said to have given a satisfactory account
of his own notion of independence. Thus Mr. Joachim apparently
maintains that truth is "independent of the intuition qua this act
of intuiting here and now." 3 Similarly, the majority of contem-
porary idealists insist strenuously that the logical or universal prin-
ciples are independent of the psychological circumstances attending
their appearance in finite minds. But so far as I know the precise

1 Proceedings of the Aristotelian Society, N. S., 1910, 10, 191, 218.
* Royce, op. cit., 118, 120, 123. According to Joachim, realism asserts that
'experiencing makes no difference to the facts' (The Nature of Truth, 33, 58).
» Op. cU., 52.

THE NOTION OF INDEPENDENCE 105

meaning of the term in this application is left to be determined by
common sense. Nor has the idealist ever given a general and pre-
cise definition of the correlative notion of dependence. This notion
is essential to the idealistic theory of the internality of relations, and
to such fundamental conceptions as 'coherence,' 'synthetic unity,'
'significant whole.' And if the idealist's polemic against realism
is successful, we are left to conclude that experiencing does "make
a difference" to facts. This is as truly the central contention of
idealism as the contrary is the central contention of realism. But
we are left in the dark as to the precise nature of the dependence
which is predicated. Idealism has even derived a certain advan-
tage from its failure to define dependence. For it has been able to
vary the meaning to suit the polemical exigency. And this ab-
sence of explicit definition has enabled idealism to profit by the
vague but natural assumption that any relation whatsoever in-
volves dependence. To prove dependence idealism has not found
it necessary to do more than to establish some sort of connection
between the term in question and some other term. The moment
dependence is distinguished from bare relation, a very consider-
able portion of idealistic reasoning is rendered worthless — a mere
recitation of the obvious and trivial.

A realist might fairly take the position, then, that he means by
independence the negative of what his opponents mean by depend-
ence. If a vague, common-sense notion will suffice in the one case,
it will suffice also in the other. The question can be argued in the
vague terms common to both parties; and this is, as a matter of
fact, what has thus far taken place. But in adopting such a course,
realism loses an important opportunity. Realism is responsible
for forcing this issue of dependence or independence, and should
undertake to clarify the conception to which it has given a fresh
prominence. Realism as a constructive doctrine is professedly con-
cerned with the merits of the question rather than with the turn of
controversy. Nor will any realist be deterred from thorough analysis
by fear of enlightening his adversary. Furthermore, the present
realistic movement is largely inspired by the logical motive, and
finds the clarification of current notions a proper and congenial task.

106 A REALISTIC THEORY OF INDEPENDENCE

II

MEANINGS OF THE TERM DEPENDENCE

THE term 'independence' is evidently used to deny 'dependence.'
Like other negative terms such as 'immaterial,' 'unworldly,' etc.,
it has acquired secondary meanings of a positive character. Thus
political independence has come to signify self-government, or
certain positive 'liberties/ such as 'free speech' and 'freedom of
the press.' Similarly, practical independence may mean com-
petency, self-reliance, or initiative. But with these derived mean-
ings we here have nothing to do. They are particular cases of
independence in which the circumstances of the application have
impregnated the general meaning of the conception. The primary
and general meaning of independence is non-dependence.

Hence we must begin our analysis with an enumeration of the
various senses in which the term 'dependence' may be intelligibly
used. One cannot be at all confident that such a list as follows is
final, either in respect of completeness or of logical coordination.
Indeed, one may feel reasonably sure that it is not. But while
inviting corrections and additions, one may hope that such a list
will at least cover the various senses of the terms that are likely to
be in question in connection with the present issue.

1. Relation. — Even though one may conclude that bare 'rela-
tion' is so radically different from the types of dependence that
follow as to justify its eventual rejection from the list, it is impor-
tant to include it provisionally. We must have this conception
before us from the outset, since it will figure so prominently in our
conclusions.

It is not possible to define 'relation.' It must either be accepted
as an ultimate logical category, or be simply cast out altogether
on the ground of the alleged dialectical difficulties in which it is
involved. But writers like Bradley who have thus rejected it have
confessed their inability to find a satisfactory substitute, and have
perforce taken refuge in agnosticism. That these dialectical diffi-

MEANINGS OF THE TERM DEPENDENCE 107

culties are artificial, has, I think, been demonstrated by James.1
All exact or analytical thinking, as at present carried on, is depend-
ent on the conception of relation; and the empirical testimony
in its favor is so overwhelming as to justify its acceptance without
further ado. It is true that any attempt to deal with relations
systematically at once encounters doubtful cases, such as 'iden-
tity' and 'difference.' But the unambiguous cases such as 'be-
fore,' 'after,' 'more,' 'less,' 'like,' 'unlike,' etc., are abundantly
sufficient to establish the genus.

2. Whole-part.2 — A whole is said to be dependent on its parts, —
on what it contains, and can be divided or analyzed into. It is
worth while to introduce at this point a distinction between 'ma-
terial' and 'formal' instances of the whole-part dependence. The
first is exhibited in the relation between the present city of London
and Trafalgar Square, or between the existing government of the
United States and President Taft. The second is exhibited in the
relation between a city and its streets, or between a government and
its chief executive. In other words, a material relation is a rela-
tion between particular values of variables, while a formal relation
subsists between the variables themselves. The dependence of
whole on part may be of either type.

3. Part-whole. — Parts are said to be dependent on the whole
to which they belong when these wholes are 'organic.' The dis-
tinction between 'formal' and 'material' may be applied here also.
Thus the hypothenuse of a right-angle triangle is formally de-
pendent on the definition of the right-angle triangle. Not only
does it derive its meaning from its participation in the whole, but
its magnitude is determined by its interrelation with other parts,
such as the opposite angle and its adjacent sides. A particular
hypothenuse is likewise both defined and determined by its ma-
terial membership in the particular triangle to which it belongs.
Similarly, an organ or member in the biological sense is said to be
dependent both formally, as respects its meaning, and materially,

1 "The Thing and its Relations," Essays in Radical Empiricism, III.

2 Cf . also Spaulding, below.

108 A REALISTIC THEORY OF INDEPENDENCE

as respects its structure and function, on the integrity of the organ-
ism to which it belongs.

But such dependence would appear to be reducible to dependence
of other types. Thus when one says that the hypothenuse depends
on the right-angle triangle for its meaning, or that the conception
of an hypothenuse depends on the conception of a right-angle tri-
angle, we are virtually naming a part for its participation in a whole.
We are virtually saying that the side-opposite-the-right-angle-of-a-
triangle cannot be such without the triangle. But this is no more
than to say that the conception of a triangle depends on the con-
ception of a triangle, which is as redundant as it is obvious. Or it
may be construed as meaning that a part cannot be a part, that is,
belong to a whole, without the whole. But this is equivalent to
saying that the complex relationship of part and whole depends on
the whole as one of its terms. And this is a case of dependence
of whole on part, and not of part on whole.

Similarly, to say that the length of the hypothenuse is materially
dependent on the magnitude of the other sides and the included
angle, is virtually to say that an interdependence of parts consti-
tutes the nature of a certain whole. The dependence of the part
is here conditional on its membership in the whole ; and its depend-
ence is on the other parts, not on the whole. We are simply say-
ing that in so far as an element belongs to a certain whole it must
possess the relations proper to it as a part of that whole. We do
not assert that the element is dependent on its membership, and
thus categorically dependent on the whole; but only that if a line
is to assume the role of an hypothenuse, it must play the part.
On the other hand, there is evidently a new kind of dependence
here exhibited by the relation between part and part. But as this
is not a part-whole dependence, it will receive consideration else-
where, under causality.

The dependence of members of a living organism may be disposed
of in the same manner.1 The respiratory system cannot be a vital
function without the whole organism. But this is merely to say

1 Cf. also Spaulding, below, 243 ff.

MEANINGS OF THE TERM DEPENDENCE 109

that it cannot belong to an organism without an organism to belong
to. To make the dependence of the part evident one must describe
the part as part-of-whole. But the dependence of member-of-
organism on organism is not a dependence of part on whole, but
rather a dependence of whole on part. It asserts the dependence
of a complex relationship on one of its terms. The dependence of
the respiratory system on the circulatory system, however, means
that the two are connected by the laws of the complex process to
which they belong; or that the one supplies the necessary condi-
tions of the other ; both of which relations would be instances of
causal rather than of part-whole dependence.

4. Thing-attribute. — Whether the thing-attribute relation is or
is not a case of the whole-part relation need not here be decided.
But it is clear, I think, that the relation presents no novelties in
connection with the matter of dependence. It is doubtful, as we
have seen, whether in some varieties of substantialism the thing
is dependent on its attributes at all. If not, then the relation is
not a case in point. But where a thing is regarded as dependent on
its attributes, it is either 'made up' of them, or defined 'in terms' of
them. It seems clear that except for an agnostic substantialism
a thing must be regarded as dependent on its attributes in that they
are in it or of it. Both would be instances of the whole-part type
of dependence, as described above.

5. Attribute-thing. — The question of the dependence of attri-
butes on the thing to which they belong, resembles the question
of the dependence of part on whole. Red cannot be attribute of
the rose without the rose ; nor would it bear the peculiar relation
that it does to odor, form, and growth of the rose, were it not for
the nature of the rose as a whole. But this will, I think, turn out
to mean either that a rose is a rose (redundancy) ; or that the red-
rose relationship depends on 'rose' as one of its terms (whole-part) ;
or that the redness of the rose is determined by its age, chemical
structure, nutrition, etc. (causation). We may therefore dispense
with the attribute-thing relation as a primary type of dependence.

6. Causation. — It is desirable so far as possible to avoid staking

110 A REALISTIC THEORY OF INDEPENDENCE

the issue of dependence on a special theory of causation. Never-
theless, it is impossible to allow a certain theory of causality to
remain at large, lest it upset our calculations at the eleventh hour.
I refer to the theory that causation is creation ex nihilo by an 'ac-
tivity.' I do not mean in the least to exclude the category of pur-
pose ; i.e. to suggest that there is no such thing as moral or rational
causation. I mean to insist only that so far as causation is observ-
able or verifiable at all, in so far as it can hi any given instance be
profitably discussed, it must be regarded as a complex or process in
which there is a relation of necessity between distinguishable and
definable parts. The cause must be displayed, as well as the effect ;
it must not be kept in the background a recondite and incalculable
factor. I shall not argue the matter further than to appeal to the
fact that the 'creation' theory has long since been discredited in
science and all other exact discourse.

If this possibility be excluded, there need, I think, be no further
occasion for dispute here. Causality is a material relation between
two complexes, derived from a primary formal relation between
their constituent variables. Thus if v=gt, for all values of these
variables, then any given velocity (v), is dependent on the constant
of gravity (gf), and some magnitude of time (f). The formal rela-
tion among the variables is called the 'law,' and the material de-
termination of the values of the variables, as prescribed by the law,
is causation.

It would, perhaps, be more in keeping with verbal usage to con-
fine the term 'causation* to a special variety of the type of de-
pendence just described; that variety, namely, in which a com-
plex occurring later in time is determined by a complex occurring
earlier in time. In other words, it is customary to limit the ad-
jective ' causal ' to laws which contain time as a variable ; and to
treat time in the positive or forward direction as the independent
variable. Or one may still further narrow the conception of cause
to mean those other values which together with time determine the
value of a future complex.

It is to be remarked that causation is conditioned by the law.

MEANINGS OF THE TERM DEPENDENCE 111

In other words, it takes place only within the system which the law
describes; and can be attributed to a complex only when the com-
plex is identified as "a case of" the system. Thus a complex which
is identified as a member of the gravitational system is caused, as
respects its position, velocity, orbit, etc., by the distances and masses
of surrounding bodies. Causes and effects are thus interdependent
within the given system, or under the law. But this leaves open the
question of whether they are dependent on the existence of the system
or the law. These determine their behavior under certain condi-
tions, but do not prove that the conditions themselves are necessary.
For it is possible that a given complex should be accounted for in
terms of one system, and yet conform to the requirements of an-
other system as well. Suppose, for example, the position of a body,
a to be defined in terms of its direction and distance from a second
body, b; and suppose it to be also defined in terms of its direction
and distance from a third body, c. It will then be the case that
the position of a is unequivocally defined in terms of either a or of 6.
Similarly, the kinetic energy of a body is definable in terms of its
equivalence to the potential energy that has been converted into
it ; or in terms of the energy of heat into which it may be con-
verted. In such cases, it is more correct to say that the complex
in question is not dependent on either determination, in view of its
possessing another determination which is sufficient to account for
it. It follows that a is dependent on b in the causal sense, only pro-
vided a is completely determined exclusively within the system in
which it is the effect of 6; only provided, in short, it has no other
sufficient cause.

7. Reciprocity. — It is customary to use the term 'reciprocity*
to express a relation of the same type as causation, but without
the same emphasis on temporal antecedence and consequence. It
is evident that the relation among the various values of the vari-
ables of a law is mutual. It is possible not only to predict the fu-
ture, but also in like manner to infer the past. Similarly it is
possible to infer simultaneities, as e.g., in the case of the configura-
tion of the planetary system, or the co-presence of extension and

112 A REALISTIC THEORY OF INDEPENDENCE

color in the visual field. It is not even necessary that time should
enter into such calculations at all ; as is illustrated by the inter-
dependence of spacial magnitudes as formulated by geometry.
'Reciprocity/ then, may be taken to mean the mutual deter-
mination of values of variables under the law, where the factor of
time-direction is not essential. Inasmuch, however, as the most
familiar cases are cases of causation, and inasmuch as the under-
lying principle is the same, I shall hereafter omit reciprocity and
speak only of causation. I shall assume, in other words, that
causal dependence is reciprocal.

8. Implying. — Finally, there is1 the simpler logical relation of
implication. It is unnecessary to discuss the question as to whether
this is or is not a ' primitive ' conception. Mr. Russell shows that
it may be expressed in terms of other conceptions, such as ' contra-
diction' and 'logical addition';1 but in any case there is some
fundamental form of logical necessity.

It is important to point out that the relation of implication is
not a symmetrical one. That which implies is dependent in one
sense ; and that which is implied, in another. Thus the premises
of a syllogism cannot both be true unless the conclusion is true ;
while the conclusion on the other hand may be true even though
the premises be false. Only the dependence of the implier on the
implied is positive and unqualified.

9. Being implied. — That which is implied, on the other hand,
is dependent on the implier only in the limited sense already noted
in the discussion of causation. For the implied may be otherwise
implied. That which is implied by two or more sets of premises
cannot be said to be dependent on any one of these sets. In the
absence of any one it would none the less be necessitated by the
others. Its dependence, in other words, is limited to the specific
logical system in question.

Because of this fact, the ordinary mathematical conceptions of
dependence and independence are not of material assistance in our
present task. The 'dependent variable' is that variable whose

1 Principia Mathematics, I, 6 ff.

MEANING OF INDEPENDENCE IN NEO-REALISM 113

value is derived by implication when a value is assigned to another
value, called the 'independent variable.' But since the operation
may be reversed, the one variable is logically as dependent as the
other. Furthermore, the question as to whether the value of the
dependent variable can be otherwise derived, is not raised. Simi-
larly, an 'independent postulate' is a postulate in a given system
that is co-determinant with the other postulates, but cannot be
deduced as a theorem from these other postulates.1 But the de-
pendence of a postulate as established by this criterion is relative
to the system in question. By virtue of being a theorem in another
system, it might be independent of the first system. We conclude,
therefore, that a is not made unqualifiedly dependent on 6 through
being implied by it, unless it is implied only by 6.

Omitting from the above list of possible meanings of the term
'dependence' those which involve needless repetition, we are left
with five : relation, whole-part, exclusive causation, implying, being
exclusively implied. It is not claimed that these are logically ul-
timate or coordinate, but only that they are intelligible, and, so
far as our main problem is concerned, complete.

Ill

THE MEANING OF INDEPENDENCE IN NEO-REALISM

WE are now in a position to formulate the realistic notion of
independence, reserving the proof and the applications for a
later portion of the paper.

1. Independence is not non-relation.2 — Realism does not deny
non-relation.3 But it is not non-relation which the realist has in

1 Cf. Huntington, E. V., Monographs on Topics of Modern Mathematics,
edited by Young, J. W. A., 169.

1 Cf. also Spaulding, below.

* Whether the conception of non-relation is tenable or not will, I think, be
found to turn upon what is made of 'difference,' 'possibility,' etc. If these be
genuine relations, then relation is universal ; otherwise not. I leave the question
open, to avoid needless complication of the issue.
I

mind when he uses the term 'independence.' Thus Mr. Joachim
is correct in supposing that according to realism, 'the facts' and
'experiencing' "are or may be related"; and that "the relation
when, or as, it obtains, leaves each (factor) precisely what it was,
viz., absolutely in itself and independent." l In other words, it is
fundamentally characteristic of neo-realism to distinguish 'rela-
tion' and 'dependence.' Otherwise, as the critics of realism have
taken pains to point out, the independence theory would be equiva-
lent to agnosticism. For if the real were necessarily out of rela-
tion to knowledge, then it is obvious that, as real, things could not
be in the relation of being known. Thus it behooves realism to
define a species of relation in which the terms, although related, are
nevertheless independent ; or to show that dependence is something
over and above bare relation.

Although realizing that Professor McGilvary is in substantial
agreement with the view here set forth, I cannot but feel that his
presentation of the matter is too easily open to misunderstanding.
"By an 'independent' object," he says, "the realist means an
object that exists when there is no awareness of it." Now this
must mean one of two things. It may mean that an object is
independent in so far as there is no awareness of it. But in this
case the only independent things are the unknown things; and
one must with Kant divide the world into known phenomena and
unknown reals. Or it may mean that an object is independent in
so far as it does not require awareness in order to exist. But this
is the same as to say that an independent object is independent of
awareness ; and we still require a definition of independence. We
require, in short, a definition of independence that shall not either
affirm or deny the fact of awareness. "If," Professor McGilvary
continues, "he (the realist) ever speaks of the qualities of which he
is aware as now being independent of awareness, he begs to be
understood as meaning by 'independent' something different from
what he means by independence when he speaks of the independ-

1 Op. tit., 41.

MEANING OF INDEPENDENCE IN NEO-REALISM 115

cnce of the qualities of which he is not aware.1 This is consistent
with Professor McGilvary's own definition of independence. But
it is clear to me that what is required for an empirical realism is a
sense of the term 'independence' that shall hold of objects equally
whether there be awareness of them or not. And such a sense must
be defined without including in the definition either the presence
or the absence of awareness.

Thus Professor Dewey is equally mistaken in supposing that
realism assumes "the ubiquity of the knowledge relation."2 Real-
ism does not argue from the 'ego-centric predicament/ i.e. from
the bare presence of the knowledge-relation in all cases of knowl-
edge. On the contrary, it denies the possibility of arguing from that
predicament at all.3 Its use of the predicament is polemical and
negative merely. It convicts idealism of so arguing, but does not
propose to fall itself into the same error. Realism defines depend-
ence as a peculiar kind of relation; so that the mere presence of
knowledge as a relation cannot be used to argue dependence. Is
being known a relation of dependence or not? If it is, then all
known things are dependent; if it is not, then things are inde-
pendent of being known, whether as a matter of fact they be known or
unknown.

2. Independence is not priority. — That which includes, implies,
causes, or explains is not independent of what is included, implied,
caused, or explained. That which is inferred or determined is not
more dependent than its premises or ground. In other words, the
difference between logical activity and passivity, or the difference
of logical direction, is not the same as the difference between inde-
pendence and dependence. Such a notion of independence appears
in all varieties of 'absolutism.' The 'ideal of reason,' converted in
more recent times into an 'ideal experience,' 'a perfect coherence/

1 E. B. McGilvary, J. of Phil., Psychol., etc., 1907, 4, 686.

* Brief Studies in Realism, II, /. of Phil., Psychol., etc., 1911, 8, 554, and
passim.

3 Cf. my article, The Ego-centric Predicament, J. of Phil., Psychol., etc., 1910,
7, 5-14.

116 A REALISTIC THEORY OF INDEPENDENCE

an 'absolutely organized experience/ or into the 'mandate' (Sol-
len) binding on the act of judgment, is conceived as the final 'pre-
supposition' of thought.1 It is regarded as independent of all
particular acts of thought in the sense that while the latter may be
psychologically antecedent, the former is logically antecedent.
The validity of the ideal is not derived from particular acts of
thought, but constitutes the standard by which the validity of the
latter is determined. Truth and being attach primarily to the
completed whole of knowledge, and to the parts or approximations
only in so far as these participate in the whole.

It is not necessary to urge what has been said concerning the
questionable character of this alleged dependence of part on whole,
or of the implied on the implier. Nor is it necessary to urge the
objection that this 'ideal of reason' upon which the whole argu-
ment turns is a meaningless combination of words.2 For the ques-
tion immediately at issue would, I think, be promptly conceded by
the idealist. The ideal whole may be prior to its parts, but these
are none the less indispensable to it. The absolutely organized
experience is made up of the finite experiences which it organizes;
the incoherences are taken over into the completely coherent whole ;
the mandatory ideal is an idealization of the judgments which pre-
suppose it. The solidarity of the whole requires that every least
part shall be and contribute precisely what it is. So that even
were it admitted that 'priority' is a sort of independence, the
ideal whole would not in the least on that account escape depend-
ence on its parts.

It is virtually the contention of idealism that the two notions
of independence just formulated are exhaustive of all the possibili-
ties. Either reality is independent of thought hi the sense of being
wholly out of relation to it, or in the sense of giving the law to it.
Thus idealism may be said to confront realism with a dilemma :
"Either your reality is" unknowable, and so utterly negligible, or it
is the ideal of knowledge itself, and so the very quintessence of

1 Cf. Joachim, op. cit.; Royce, op. cit. ; Rickert, H., Der Gegenstand der Erkenntnis.
» Cf. the author's Present Philosophical Tendencies, Ch. VIII, §§ 7, 8.

MEANING OF INDEPENDENCE IN NEO-REALISM 117

thought." It is plain that the whole case for realism must rest on
the assertion of a third alternative. It must be possible to regard
reality as sustaining, or as being capable of sustaining, the relation
which constitutes knowledge, while at the same time sustaining
that relation only accidentally. And if reality be "the ideal of
thought," as in a sense it undoubtedly is, then it must on realistic
grounds be possible to regard this as a role which reality assumes
without prejudice to its independence. It will not be sufficient to
assert that reality is 'prior' to finite thought in the sense that finite
thought is regulated or determined by it; it must be further as-
serted that this very regulation or determination is gratuitous, so
far as reality is concerned. It must be shown that though reality
be related to thought as its ideal, or presupposition, that relation is
of the non-dependent type. It is clear, in short, that another mean-
ing of independence is called into play, and that this third meaning
is crucial.

3. Independence is the total absence of dependence in the senses
enumerated above. — In order to prove the dependence of a on 6
it is necessary to show that a contains 6 ; or that a is the cause
or effect of 6 in a system which exclusively determines a ; or that
a implies b ; or that a is implied exclusively by 6. To exhibit any
relation of a to 6 other than these is beside the point. Whether
a and b be otherwise related, or not, does not affect the independ-
ence of a. And if it can be shown that a and b are related, and yet
not dependent in any of these senses, the relation in question is by
definition a non-dependent relation.

This is a suitable occasion on which to eliminate three current
misconceptions.

A. In the first place, the realist does not propose to define reality
in terms of its independence.1 This would be a palpable and
clumsy self-contradiction. If a is independent of 6, then a must
be definable, if at all, in terms other than b. Independence itself
is not a relation, but the absence of a certain type of relation.
Hence independence itself does not define anything. If a be re-

1 Cf. Royce, op. tit., 66, 92, 93, 108.

118 A REALISTIC THEORY OF INDEPENDENCE

lated to b, and yet independent, this is equivalent to saying that a
can be defined without reference to this relation.

B. In the second place, realism does not assert that "everything
that is true of" a, is independent of b.1 For a's independence of 6
is true of a; and this judgment evidently depends on b. The
independence of a as respects 6 expressly means that the 6-things
that may or may not be true of a are in any case not necessary to a.
The doctrine turns entirely on the distinction between what a
depends on, and what is merely true of it.

C. Thirdly, realism does not deny that when a enters into a
relation, such as knowledge, of which it is independent, a now
acquires that relation, and is accordingly different by so much;
but denies only that this added relation is necessary to a as already
constituted. Thus when a is known, it is a itself, as constituted
without knowledge, that is independent of that circumstance.
The new complex known-a is of course dependent on knowledge
as one of its parts.

IV

A REALISTIC THEORY OF INDEPENDENCE FORMULATED IN GENERAL

TERMS

WE are now in a position to advance the notion of independence
as a theory : in other words, to set forth its reasons and its applica-
tions.

1. All simple entities are mutually independent. — Simple enti-
ties cannot be dependent2 in the whole-part sense because as simple
they cannot be wholes composed of parts. Simple entities cannot
be causally related because they cannot be values of variables,
since this again would belie their simplicity. And it is acknowl-
edged by all logicians that simple entities can neither imply nor
be implied, these being relations confined to propositions or com-
binations of propositions.3

1 Royce, op. cit., 117.

2 The term ' dependence ' will henceforth be employed in the sense defined
above, 113. s Cf . Russell, B., Principles of Mathematics, 14, 15.

INDEPENDENCE FORMULATED 119

2. Simple entities are independent of the complexes of which they
are members. — It is evident that a complex cannot be a part of one
of its own components. Nor can a simple constituent sustain
relations of either causation or implication with its including com-
plex.

3. Complexes are mutually independent as respects their simple
constituents. — This follows from the previous assertion. If the
constituents into which a complex may be analyzed do not depend
on that complex, the complex itself may be destroyed without
affecting the constituents. Therefore two complexes having some
constituents in common are not made interdependent by that fact.
Consider the argument advanced by Professor Royce. He sup-
poses two independent entities to have some quality in common,
such as 'redness' or 'roundness.' One of the beings is then sup-
posed to be destroyed; while the other, being independent, sur-
vives. But if the first being is destroyed, 'redness' must go with
it; hence the surviving being cannot possess 'redness/ which
contradicts the original supposition. "It follows," he concludes,
"that the many entities of the realistic world have no features in
common." x But the argument turns entirely upon the assump-
tion that when an entity is destroyed its qualities are destroyed
likewise, or that the simple constituents of a complex are de-
pendent on the complex ; and this assumption, as we have seen, is
false.

4. Complexes as wholes are dependent on their simple constituents.
— The cherry is dependent on 'redness/ 'roundness/ etc. This
is no more than a restatement of one of our definitions of depend-
ence. But when taken together with the previous assertion it
reveals the important fact that dependence is not always recipro-
cal. While a complex depends on the terms into which it may be
analyzed, these are none the less independent of the complex.

5. A first complex is dependent on a second complex when the
second complex is a part of the first. — This also is simply the re-
statement of the whole-part type of dependence. But it is im-

i Op. cit., 130, 131 ; cf. 114.

120 A REALISTIC THEORY OF INDEPENDENCE

portant to observe that here again dependence is not reciprocal.
A complex part is not dependent on its including whole simply by
virtue of its participation therein ; but only in so far as it stands
in relations of dependence with the other parts. The members
of a collection are not dependent on the collection, but may be
dependent by causation or implication on other members of the
collection. The ordinary supposition to the contrary is due to a
confusion that virtually begs the question. Thus we may say
that the members of the planetary system depend on the whole
system for their being members-of-the-planetary-system. But
this is true only in the trivial and redundant sense. It does not
prove that Jupiter, e.g., is dependent on the collective planetary
system, which is the very question at issue. Jupiter is dependent,
however, not on the planetary system as a whole, but on the sun,
Saturn, etc., as causes and effects, or on the law of gravitation as a
premise of implication.

It follows that when two complexes are interdependent, this
does not involve the interdependence of their parts. Since a does
not depend on abc, a is not necessarily dependent on r, even when
abc is dependent on rst.

6. A first complex is dependent on a second complex when the first
is either cause or effect of the second within a system which exclusively
determines the first. — Thus Jupiter is dependent on the sun inas-
much as its velocity is a function of the sun's mass according to a
law which alone accounts for that velocity. If the velocity of Jupi-
ter were deducible from the plan of God regardless of the mass of
the sun, then despite its conformity to the law of gravitation it
would be independent of the sun. Or, if one preferred, one might
say that it would then be dependent on the sun within the planetary
system of gravitation; it being understood that it would be inde-
pendent of that system by virtue of its place in the plan of God.
Similarly, the mass of Jupiter cannot be said to be dependent on
the mass of the sun, inasmuch as it is definable in terms of its own
satellites. Or it could be said to have a conditional dependence
on the sun's mass, relative to its gravitational relations with the sun.

INDEPENDENCE FORMULATED 121

But even supposing the velocity of Jupiter to be determinable
only in the planetary gravitational system, it is important to re-
mark that it is the velocity of Jupiter, or some such gravitational
property, that is then dependent. If one wishes to say loosely
that "Jupiter is dependent," then one must recognize that this is
so only in so far as Jupiter is dependent on its gravitational prop-
erties as its parts. Dependence by causation is reciprocal; but
where it is complicated with the whole-part relation the resulting
dependence is not necessarily reciprocal. Thus while Jupiter as a
whole is dependent on the motion of the sun by virtue of compris-
ing gravitational properties that are causally dependent thereon,
it does not follow that the motion of the sun is dependent on
Jupiter as a whole, although it is dependent on its gravitational
properties. Being independent, let us say, of the apparent color
of Jupiter when seen from the earth, it is then independent of the
whole Jupiter when this is taken to comprise that color.

7. A first complex is dependent on a second complex when the first
implies the second. — Thus the premises of a syllogism depend on
the conclusion, and the law of a mathematical or physical system
together with the values required for the solution of the equation,
depend on the value of the unknown quantity. It is important to
remark that it is that which implies that is dependent, and not its
components taken severally. The major premise of a syllogism
does not depend on the conclusion, nor the law on a particular
cause or effect ; for the single premise, or the bare law, do not of
themselves imply. So that the falsity of the conclusion does not
necessarily disprove the major premise but only the combination
of premises ; nor does the non-occurrence of an effect disprove the
law, but only the occurrence of the cause under the law.

8. A first complex is dependent on a second complex when the first
is implied by the second, and is not otherwise implied. — If a conclu-
sion follows from several alternative pairs of premises it cannot
be said to be dependent on any one pair. But if a certain pair of
premises constitute its sole determination, then it belongs to them,
and is dependent on them. Here, again, dependence is not neces-

122 A REALISTIC THEORY OF INDEPENDENCE

sarily reciprocal. In other words, that which is implied may be
independent of the implier, despite the fact that the implier is de-
pendent on the implied.

We have thus discovered several instances of non-reciprocal de-
pendence; and in so doing have removed a dialectical objection
that has been urged against realism with some force. If 'objects'
be independent of the 'ideas/ it does not follow from the formal
notion of independence that 'ideas' must be independent of their
'objects.' 1 For the relation in question may belong to any one of
the several types of non-reciprocal dependence described above in
(2), (4), (5), (6), and (8).

9. A first complex is independent of a second complex whenever
the first is not dependent on the second in any of the senses enumerated
above, regardless of their being otherwise related.

In other words, it is not necessary to present a list of non-depend-
ent relations. Independence is not a question of relation or non-
relation, but of the presence or absence in any given case of a cer-
tain type of relationship. Entities are independent unless they are
proved dependent. Their bare relation is in the great majority
of cases discovered before any dependence is proved ; and in in-
numerable instances no such dependence is proved at all. Things
are ' together ' in consciousness, or in space, they ' succeed ' one an-
other in time, they are 'different,' 'more,' 'less,' whether or not
they are whole and part, cause and effect, or implier and implied.
These simpler relations are entirely intelligible ; and must be so
regarded even by the most extreme advocates of interdependence.
For they enter into all cases of dependence. Such relations hold,
for example, of the several postulates, constants, values, etc., of a
deductive system, and of the parts of an organic unity. It is im-
possible to reduce relation to dependence, to reduce temporal suc-

1 Cf. Royce, op. cit., 119, 69. Realists themselves have cited consciousness
as a case of non-reciprocal dependence, but without, so far as I know, discussing
the matter in detail. Cf. Russell, B., Meinong's Theory of Complexes and
Assumptions, Mind. N. S. (1904), 13, 515.

INDEPENDENCE FORMULATED 123

cession to physical causation, for example ; because the notion of
physical causation is more complex and includes the notion of
temporal succession. A complex notion can be no clearer than
the simpler notions that enter into it. Dependence is a complex
notion that is intelligible only provided the simpler notion of rela-
tion is intelligible. If bare relation be a "miracle," x then depend-
ence is a compound miracle.

There can be no logical presumption in favor of dependence.
Because things once thought independent are afterwards dis-
covered to be dependent, we may distrust our present judgments
of independence ; but if so we are governed by psychological and
not by logical motives. There is as much ground for the plain
man's feeling of wonder at the laws of nature, as for the idealist's
grieved surprise when his attention is invited to an external rela-
tion. In other words, there is no logical ground for either emotion.
If one is used to employing the method of inference, one is shocked
by unmitigated facts ; if one is used to aggregates, sequences, and
contrasts, one is startled to discover identities and widely ramify-
ing necessities. But there is no logic that has ever been con-
ceived that prefers the one to the other. The assumption of
dependence where it is not found is not only a dogma; it is a
superstition that none' of its devotees have ever subjected to a
searching examination. Had they done so, they would have
been forced to the conclusion that a pure, or entire, dependence, is
a meaningless combination of words.

The question of independence, then, is an empirical question
that must be raised over again for every case under dispute.
Given an entity a, and a second entity 6, one must inquire whether
6 is a part of a, or whether a implies 6, or whether a is exclusively
determined by a system in which it is cause, effect, or implication
of 6. An affirmative answer to any of these questions asserts the
dependence of a on 6. A negative answer to all of these questions
is equivalent to the assertion of the independence of a on 6. And

1 Cf. Joachim, op. cit., 44, 49.

124 A REALISTIC THEORY OF INDEPENDENCE

if such a negative answer is reached, then such relations as a
does sustain to 6 cannot prejudice its independence.

10. A first entity may acquire dependence on a second entity. —
This statement must be carefully guarded, and is true only in a
very limited sense.

A. In the first place, a simple entity, a, may enter into a com-
plex which is dependent on a second complex containing a simple
entity 6, of which a was formerly independent. Thus a may
enter the complex aim which is the cause of brs; as when the
round sun is the cause of the red sunset, 'roundness' having been
independent of 'redness.' But, as we have seen above (§ 3) the
dependence of complexes does not involve the dependence of their
simple components. These are as independent as they were
before, despite their figuring in an instance of dependence. This
conclusion is evident unless a is identified with aim, and 6 with
brs, which is a contradiction in terms.

B. In the second place, a complex may become dependent on a
second complex of which it was formerly independent. Thus a
body may move into a new field of force and so acquire a causal
dependence de novo. This affords a proper instance of acquired
dependence, provided it be admitted that the body in question
has changed. The motions of a body a up to a certain time were
dependent on certain bodies b, c, etc., within the field of force M ;
and independent of certain other bodies, r, s, etc., lying in a
second field of force N. After that time a is dependent on r, s,
etc. In other words, a has changed from a1, which is independent,
to a2 which is dependent on r, s, etc. But it may be objected
that since a1 and a2 are reciprocally dependent, therefore a1 is in-
directly dependent on r, s, etc., and a2 on 6, c, etc. In other
words, the body's moving into the field of force N is a function of
its determination by the field of force M; and vice versa. We
must not, however, hastily conclude that two things dependent on
the same thing are dependent on each other. For as we have
already seen, causal determination in order to involve depend-
ence must be exclusive. And a at the moment of passing into

INDEPENDENCE FORMULATED 125

the second field of force is determined by both; i.e. its posi-
tion, velocity, etc., could be accounted for in terms of either
system. Hence if regarded as member of one, it is independent
of the other. If, then, we indicate by a1 the body a so far
as wholly determined by system M, and by a2 the body
so far as determined by system N, we may say that a1 is inde-
pendent of r, s, etc., the members of system N; and that a, in
changing from a1 to a2, becomes dependent.

It is impossible to say that an entity can acquire dependence
only when we include in our definition of the entity in question,
all that ever happens, or may possibly happen, to it. This is
what Professor Royce, e.g., expressly does. "The man in China
who may become my enemy or my neighbor," he says, "is already
such that certain changes in him, if they occurred, would not be
indifferent to me. This possibility already makes part of his
being." 1 Now the possibility in question can be construed in
either one of two ways. In the first place, it may be construed
as the present actual nature of the man in China. But this is
indifferent to me. In the second place, it may be construed as
the man's present actual nature together with a hypothetical rela-
tion to me. This, it is true, is not indifferent to me, but only
because I have included the difference to me in the hypothesis. It
in no way establishes my dependence on the man in China as
presently constituted. And even if we grant the eventual depend-
ence, even that does not prejudice our present independence.
For even if our paths do cross, the point of intersection is deducible
from the antecedents in my own life history, quite regardless of
the earlier days of the man in China. I should reach that point
anyway, following my own course, so that my susceptibility to his
influence, my coming within his range, is not dependent on the
earlier stages of his course.

Furthermore, it is pertinent to observe, if I decline to define any-
thing short of all that does or may happen to it, I must not only

1 Op. cit., 126.

126 A REALISTIC THEORY OF INDEPENDENCE

deny change from independence to dependence; I must deny
change altogether. For as fast as I attribute change to any entity
a, that change is taken up into its nature ; and a thus qualified does
not change. And the same will hold true of difference, or of any
other relation or predication whatsoever. But it is needless to
press this objection. It is sufficient for the purpose of a realistic
philosophy to say that if there can be change at all, there can be
change from a state of independence to one of dependence.

A REALISTIC THEORY OF INDEPENDENCE APPLIED TO THE CASE
OF KNOWLEDGE

1. When an entity is known or otherwise experienced it is related
to a complex. — It is impossible to furnish a justification of this
assertion without undertaking a complete account of the nature of
consciousness. But it is desirable so to explain it as to reduce
opposition to a minimum. This assertion would contradict the
supposition that in consciousness there is no difference between
subject and object, or between consciousness as agent and conscious-
ness as content. It would also contradict the supposition that the
subject or agent in consciousness is a simple 'activity' or 'sub-
stance.' On the other hand, it would agree with a theory which
regarded the subject of consciousness as a context into which the
object is brought by virtue of a peculiar relation ; or with a theory
that regarded the subject as an 'apperceiving mass/ or background
of feeling, or organized self-consciousness, to which the object
known or experienced is assimilated; or, finally, it would agree
with the view that the subject in consciousness is the living and
responding organism.1

2. Simple entities are not dependent on consciousness. — There is,
as we have seen, no sense in which simple entities can be said to be
dependent at all. It follows that in so far as the knowledge of such

1 Cf. my Present Philosophical Tendencies, Ch. XII.

THE CASE OF KNOWLEDGE 127

entities is possible, they must be regarded as independent of knowl-
edge. But the knowledge of such entities is involved in the method
of analysis. If one is to recognize a complex as such, one must be
able to ascertain its simple components ; for a complex depends on
such simple components both for its nature and its meaning.

It is not necessary to assert that simple entities can ever stand
alone in knowledge, that they can be known without knowing some-
thing else at the same time. It may well be that they must be
known together with some context or schematism. There may be
a minimum cognosdble, which is a complex. But this does not
affect the question whether simple entities can be known. That
such is the case is the universal testimony of analysis. Empiri-
cism claims to know simple 'sensory' qualities, or 'impressions.'
Rationalism claims to know logical ' indefinables ' or 'categories.'
And we must conclude that in so far as such elements are known
they furnish instances of independence.

Nor, indeed, can a philosophy which rejects analysis avoid the
same conclusion. Such a philosophy merely differs from other
philosophies in respect of what it holds to be simple. Whereas
the devotees of analysis regard 'self,' 'activity,' 'substance,' etc.,
as complexes, this philosophy declares them to be indivisible. We
may fairly inquire, then, for the sense in which such indivisibles
are to be regarded as dependent. They cannot be wholes depend-
ent on parts; they cannot cause or be caused in the scientific
sense ; they can neither imply nor be implied. It would, then, be
meaningless to speak of them as dependent. If they are none the
less denied independence of knowledge, then they must be regarded
as identical with knowledge. No realist or other sane person would,
of course, propose to regard a thing as independent of itself. But
one who denies realism on such grounds must be prepared to deny
the difference between object and subject of knowledge, and iden-
tify being altogether with the act of knowing. Such a view does
not require attention until some serious effort has been made to
answer the objections that have long since been urged against it.

There is an interesting corollary to the conclusion we have al-

128 A REALISTIC THEORY OF INDEPENDENCE

ready reached. If simples are independent, it follows that knowl-
edge escapes subjectivity in proportion as it carries analysis
through to the end. The ultimate terms of experience are at any
rate independent, whatever may turn out to be the case with cer-
tain complexes of these terms. If the knower desires to eliminate
the personal equation and seize on thing-in-themselves, his safest
course is to sift experience to its elements and thus obtain a sure
footing in the independent world. Such elements, whether sen-
sory qualities or logical indefinables, will afford him a nucleus of
independence to which he may add such complexes as will satisfy
his criterion.

The present is a suitable occasion on which to comment on a
sentiment with which such a view as the above has had to contend.
Thomas Reid referred to Hume's abolition of substance as a turn-
ing of the elements of experience "out of house and home . . .
without friend or connection, without a rag to cover their naked-
ness." 1 'Tender-minded' idealists have been moved by the same
sentiment — a sort of vicarious nostalgia. "'Greenness,' 'Har-
mony,' 'Equality,'" says Mr. Joachim, "are to remain eternally
and unalterably themselves, whether they are also experienced or
not. They are 'the facts,' and they are there independently and in
themselves. But what is their being there?"2 If this argument
has any weight, it is derived from a careless use of pronouns. The
hard-hearted realist is quite ready to conclude that the simple
elements are nowhere. They may enter into this or that group,
but they do not belong to it ; they have no home. The benevolent
idealist, on the other hand, offers 'experience' or 'consciousness'
as a public refuge for all ontological outcasts. It is the same senti-
ment that inspires the belief that there must be some last defini-
tive word that can be 'said' about everything. Reality must be
defined ; everything must be brought into the fold lest it perish in
outer darkness. It is important to devise something that can be
said of everything ; and you can say of everything that it either is

1 Inquiry into the Human Mind (1895), 103.
1 Op. tit.. 40.

THE CASE OF KNOWLEDGE 129

experienced or "would be present to a sort of experience which we
ideally define." l So idealism derives a certain support from the
sentimental demand or supposed logical need of some envelop-
ing characterization of things, of some permanent address where
things may be always reached despite their wanderings. With the
sentimental demand we need not trouble ourselves, while the logi-
cal need is the very question at issue. Is there one relation of
dependence which all things sustain, or not ? If we avoid begging
the question, and are critical in our use of the term 'dependence,'
we must, I believe, conclude as above, that simple elements, at
least, depend on no relation. They are the entities 'at large/ and
belong exclusively to no constituency.

3. Complexes are independent of knowledge as respects their simple
constituents. — This is a further corollary of the conclusion reached
above, and requires to be stated separately only in the interests of
clearness. Whatever conclusion may be reached as to the de-
pendence of some complexes on knowledge, it is important to ob-
serve that this can in no way prejudice the independence of the
terms into which they can be analyzed. If we should conclude, for
example, that an imaginary complex is dependent on the act of
imagination, it will none the less remain true that such elements as
'blue' or 'identity,' if they be found in the complex, are independent
of the imagination. In other words, such dependence as there is
must attach to complexes as such, and cannot involve their ulti-
mate parts.

4. The propositions of logic 2 and mathematics are independent
of consciousness. — We have now to do with the independence of
some complexes, assuming that each type of complex must be dealt
with on its merits. We have only to select an instance of the type
and apply the criteria already adopted. Is the proposition,
c2 = a2+62— 2 ab • cos y, where y is the angle of a triangle, c the op-
posite side, and a, b the adjacent sides, dependent on the rela-
tion to knowledge?

1 Royce, Conception of God, 30.

2 Cf . Marvin's proof that knowledge 'presupposes' logic, above, 51 ff.

K

130 A REALISTIC THEORY OF INDEPENDENCE

In the first place, the above proposition does not contain the rela-
tion of knowledge, as one of its parts, as it does contain 'line/
'equality,' 'angle,' etc. The relation to knowledge is not to be
found in it by analysis. This is true of all the notions of the cog-
nitive relation mentioned above (I).1 There is no relation to a
background of feeling, or to an apperceiving mass, or to the activ-
ity of a self or responding organism. The proposition in question
is therefore not dependent on knowledge in the whole-part sense.
Nor does the proposition imply any of these relations. The only
serious question is whether it is causally determined, or implied by
such a relation. But as we have seen, this is not itself decisive as
respects its dependence or independence (see II, 6, 9). Assuming
for the moment that the proposition is implied by knowledge, and
does sustain causal relations with the subject of knowledge; we
have still to inquire whether it is thus exclusively determined.
And it is evident that this question must be answered in the nega-
tive. For the proposition is sufficiently determined, without refer-
ence to knowledge, by the logical and mathematical systems to
which it belongs. It is implied by a set of postulates, and is cause
and effect in relation to coordinate theorems. In other words,
whether it be determined in the knowledge relation or not, it is
in any case not so determined exclusively. This may be expressed
more loosely by saying that even were it not necessary for cognitive
reasons, it would still be necessary for logical and mathematical
reasons ; so that its cognitive necessity does not make it dependent.

Thus the proposition in question, since it is not dependent on
the knowledge-relation in any of the accepted senses, may be de-
clared to be independent thereof.

5. Physical complexes are independent of consciousness. — The
question of the independence of physical nature introduces no
novelties. The mean velocity of the planet Jupiter, for example,

1 The argument would be more empirical and decisive were I to employ only
what I regarded as the true conception of the cognitive relation. But as such a
course would narrow the scope of our conclusions I have so far as possible left the
question open.

THE CASE OF KNOWLEDGE 131

neither contains nor implies the cognitive relation. Assuming
that this complex is implied by the knowledge of it, and that it
sustains causal relations with the subject of knowledge, it is none
the less independent because of the fact that it is completely de-
termined by other relations, such as its distance from the mass of
the sun. It can be deduced, and has, as a matter of fact, been de-
duced, from the celestial gravitational system without reference
to cognition.

But I wish in this connection to profit by the powerful support of
Mr. L. T. Hobhouse. In his "Theory of Knowledge," this writer
declares that the "independent existence" of A is a "negative
characteristic of A's existence." "It says," he continues, " 'The
A which I now apprehend would exist now and would still be A
even though I did not apprehend it, and thus (for example) it
may continue to exist, though I should cease to apprehend it. ...
We are brought, then, at once to the question, How can this
independence be known ? And the answer is, that it depends en-
tirely on our success in discovering universal laws in the occurrences
of phenomena.'"1 He concludes, in other words, that where a
physical event can be inferred from other physical events by
virtue of an established law, the inferred event can be regarded as
independent of other conditions, such as its "apprehension," that
are not required for its inference.

But Mr. Hobhouse proposes a method of eliminating appre-
hension altogether. Thus if B, which is known from observation
to be the effect of A, is given when A is not apprehended, we may
infer A to be causally operative despite its not being apprehended. In
other words, we may now conclude that although when the law was
discovered A was apprehended, its being apprehended was not a
condition of its effectiveness. This is, perhaps, the most convinc-
ing way of presenting the argument. Nevertheless, it is impor-
tant to note that it involves a dangerous and needless concession
to the opponent. For it is not necessary to eliminate a condi-

1 Hobhouse, L. T., Theory of Knowledge, 522 (italics mine).

132 A REALISTIC THEORY OF INDEPENDENCE

tion in order to disprove its necessity. If A can be shown to be
the cause of B, so that B can be inferred from A alone, this is suffi-
cient to prove the independence of B on C, whether C as a matter
of fact happens to be present or not. B is dependent only on those
parts of the context which exert determination upon it, or re-
quire to be employed in deducing it. Strictly speaking, it is never
possible to obtain an empirical instance in which only the deter-
mining conditions are present. It is the task of science to distin-
guish within a total manifold those factors which do count and
those which do not. Thus the determination of the length of a
side of a triangle by a specific ratio of the magnitudes of the opposite
angle and its adjacent sides, is discovered within a fuller context,
containing, for example, the absolute magnitudes of the adjacent
sides. And at the same time that it is discovered that the ratio
in question does count, it is found that the absolute magnitudes do
not count. Similarly, when Galileo discovered that acceleration
was a function of the time of a body's fall, he discovered that it was
not a function of the body's weight or volume. And to establish
this it was not necessary for him to obtain an instance of a body
without weight or volume ; it was sufficient for him to show that
the factors, although present, did not enter into the calculation.

We may conclude, therefore, that in so far as physical phenomena
are deducible from physical causes without reference to conscious-
ness, they are independent of consciousness, even though conscious-
ness be present ; even, indeed, though they were to prove deducible
from consciousness also. In short, if physical event B be suffi-
ciently determined by physical cause A, B is independent of C,
whatever its relation thereto.

6. Logical, mathematical, and physical complexes may be or be-
come objects of consciousness, despite their independence. — Since
dependence has not been identified with bare relation, the assertion
of independence does not involve the assertion of non-relation. If
knowledge be defined merely as a relation, it is, therefore, impos-
sible to argue that a thing's independence forbids its being known.

The question cannot, however, profitably be discussed in terms

THE CASE OF KNOWLEDGE 133

so general and non-committal. We must suppose that when a
thing is known, it enters into a system which is internally deter-
mined. If we are not to regard the subject of knowledge as a
simple indefinable, and its relation to its objects as an ultimate
relation of which no more can be asserted than its disjunctive or
external character,1 there is only one course open to us. We
must observe the knowledge process in the concrete, and take
into account whatever physiological, psychological, or ethical fac-
tors it appears to involve. If such a course is adopted, we can
scarcely deny that the knowledge process has laws of its own;

»and that the parts of the knowledge process, including the object,
must come under the terms of the law and be determined by it.
In other words, if a thing is known, it must submit to the con-
ditions which knowledge imposes.

Thus Mr. Joachim says, paraphrasing what he regards as the
independence theory: "Greenness is an entity in itself. And
though, as experienced, it is related to a sentient consciousness,
yet even in that relation it remains in itself and unaffected by the
sentience." He then very fairly inquires, "Is it then irrelevant
to the nature of greenness what the nature of the sentience may
be? Clearly, the sentience to which greenness can be related is
'vision/ not 'hearing.' But we are to understand that this re-
striction is not based on the nature of greenness as such, but is
just a fact. And presumably also the restriction in the range of
the sentience — the restriction, e.g., of vision to color, of hearing
to sound, of this type of vision to greenness, etc. — is just a fact,
which in no way enters into the nature of the sentience." 2

I regard Mr. Joachim's remarks as entirely pertinent. The
relation between greenness and vision is not arbitrary. To sup-
pose so would be to ignore certain well established conclusions of
physics and physiology. Nor is the relation of the object to the

1 This appears to be the course adopted by Messrs. B. Russell and G. E. Moore.
To the present writer it seems to over-simplify the issue, and avoid very pertinent
questions. Cf. Moore, The Refutation of Idealism, Mind, N. S. (1903), 12, 442,
449, 453. 2 Op. cit., 43 (italics mine).

134 A REALISTIC THEORY OF INDEPENDENCE

other factors of consciousness, within the conscious process itself,
ever arbitrary. At any rate there is as much evidence of law
and determination here as anywhere else.

In discussing this matter further, I shall employ that doctrine
concerning consciousness which I personally hold to be true. In
this case, I feel justified in doing so because the doctrine in ques-
tion does not evade Mr. Joachim's difficulty, but exhibits it in
the strongest possible form. Let us assume, then, that con-
sciousness is a process containing a nervously endowed organism,
a specific type of response to stimulation, and portions of an
environment selected by the response. Let us assume, further-
more, that this operation as a whole is interested or teleological.
It follows that when there is consciousness of B, B is introduced
into a system governed by two types of law. On the one hand,
B will now obey the laws of optics, acoustics, etc., determining the
interrelations of physical stimuli and physiological sensory mech-
anisms. On the other hand, B will now obey the biological and
ethical laws which govern the action of an organism on its en-
vironment. B, in so far as known, is determined by the subject
of consciousness, whether this be regarded as a physiological com-
plex or as a moral agent. And we must conclude that B is there-
fore deducible from these factors under the terms of the laws
governing its relations with them. So that did we but know our
sense-physiology, our biology, and our ethics, as well as we know
our celestial mechanics, we could presumably deduce B from our
consciousness of it; or 'greenness,' e.g., from the sensory process
by which it is apprehended.1

Are we, then, to conclude that 'greenness' is dependent on the
sensation of it ? No ; and for a reason that has already been set
forth. To prove B to be dependent on C it is not sufficient to
prove that B is implied or causally determined by C. It is neces-
sary to prove that it is exclusively determined by C. And in so

1 In order to make such a deduction it would, of course, be necessary to possess
laws governing the interaction of organism and environment. Laws governing the
action of the organism by itself would not suffice.

THE CASE OF KNOWLEDGE 135

far as B is implied or causally determined by A, as in the case of
physical events, that determination establishes B's independence.
We may conclude, in other words, that even though an object
enter into relations of determination when it is known, such rela-
tions do not prejudice such independence as it possesses by virtue
of its logical, mathematical, or physical determination. In so far
as any given object is deducible otherwise than from conscious-
ness, it is independent of consciousness.1

The realist is by no means one of "those who admit that the sole
and exhaustive relation of the 'self or 'ego' to objects is that of
knower of them." He would willingly grant that "one who is
knower is, in relation to objects, something else and more than
their knower," and that "objects are, in relation to the one who
knows them, something else and other than things in a knowledge
relation." 2 He does not, in other words, deal with the knowing
relation abstractly, but regards it as a complex process, involving
physical, physiological, biological, and ethical factors that are de-
terminable by the laws proper to these sciences. I do not wish to
limit the extent to which this determination may go,3 nor even
to preclude the possibility of there being physical complexes ex-
clusively determined by the organic processes involved in conscious-
ness. But even in such cases the principle of independence would
not be endangered. If perceiving modifies its objects as one body
modifies another, then we must attribute to the object at least as
much independence of consciousness as we attribute to one body
in relation to another. No body is ever wholly dependent on
another body. Its being modified by another body means that the
second body makes some difference — but not "all the difference."
Hence the alteration of the perceived body by the physiological
mechanism of perception could in any case prove only that a body

1 Cf. Pitkin, below, 396, etc., for evidence showing that on the whole it is char-
acteristic of organic response not to disturb the environment ; that is, not to intro-
duce into it factors which are not determined by its own non-biological laws.

*Dewey, J., Brief Studies in Realism, II, /. of Phil, Psychol, etc., 1911, 8,
651, 552.

3 Cf. Pitkin, below, 405.

136 A REALISTIC THEORY OF INDEPENDENCE

otherwise independent of consciousness is in a certain limited respect
dependent thereon. Furthermore, considerations such as these
could not possibly be used to establish either the universality or
the uniqueness of dependence on consciousness. It would be a
dependence happening to some bodies, under peculiar conditions
which sometimes obtain, and more often do not. It would be only
a special case of a very common kind of dependence, arising from
the fact that the organism belongs to the field of interacting bodies.
Thus a thing may become known and cease to be known (IV,
10), even though knowledge involve a modicum of dependence.
Thus B, e.g., might, owing to physical reasons alone, be brought
within the conscious process. Its earlier history, including arrival
at the point of entrance into consciousness, would be determined
by its physical antecedents. From thenceforth its history would
be determined by new laws, remaining independent of them just
in proportion as it could still be accounted for in terms of the
old laws. It might then drop out of the new system and there-
after be exclusively determined by its physical conditions. In
this way a body might have a continuous history that is independ-
ent of consciousness, despite an interval of determination by
consciousness.

VI

CASES OF SUBJECTIVITY, OB DEPENDENCE ON A PRIMARY CONSCIOUS-
NESS

BY 'subjective' I shall mean whatever is dependent on conscious-
ness. It is important to recognize the existence of such a cate-
gory, and to present instances that will illustrate the meaning of
independence by contrast. We may also hope in this way to de-
velop a power of discriminating doubtful cases; although many
cases must for the present remain doubtful.

We must distinguish at this point between dependence on a
primary and on a secondary consciousness. We shall find that a

DEPENDENCE ON A PRIMARY CONSCIOUSNESS 137

thing which is dependent on one consciousness-relation is inde-
pendent of others ; or that cases of subjectivity are themselves in
an important sense independent.

1. Parts of consciousness as such, are dependent on the whole of
consciousness. — An object-of -consciousness cannot be such with-
out consciousness. As we have seen (III, 3) this is really equiva-
lent to asserting the dependence of a whole on its parts ; it asserts
the dependence of consciousness in a broader sense, on conscious-
ness in a narrower and included sense. As it asserts the de-
pendence of consciousness itself, and not the dependence on some-
thing else on consciousness, it is included here, therefore, only for
the sake of clearness.

But it is worth while again to emphasize the asymmetrical charac-
ter of whole-part dependence in this application. Mr. Russell
observes that awareness is "utterly unlike other relations, except
that of whole and part, in that one of its terms presupposes the
other. A presentation must have an object." l Were Mr. Russell
to enter into the particulars of the question, he would find, I think,
that presentation is a case of whole and part; and that the one-
sided dependence of presentation on object is only a special case
of the asymmetrical dependence of whole and part.

It is further to be observed that even should a complex prove to
be dependent on the subject of consciousness, this would in no
way involve the dependence of its parts on the subject of conscious-
ness, whether those parts be simple or complex. Still less would it
involve the dependence of such parts on the parts of the subject.

2. Parts of consciousness are reciprocally dependent within the
system of consciousness, but only in a limited sense. — In other
words, a thing is a part of consciousness by virtue of the action of
the other parts ; and once it is such its behavior is conformable to
the laws of consciousness. Thus when I perceive B, B is depend-
ent for the status 'perceived-object/ upon the act of perception.
And in the new r61e of the 'perceived-object/ it obeys the laws of

1 Russell, B., Meinong's ' Theory of Assumptions and Complexes,' Mind, N. S.,
(1904), 13, 515.

138 A REALISTIC THEORY OF INDEPENDENCE

perception, and is determined by the other factors involved in
perception. Thus if we limit our view to the system objects-
perceived-by-M , B's history therein, its appearance, alterations, and
disappearance, are functions of the subject M. But, as we have
seen, this does not mean that B is unqualifiedly dependent on M,
unless it has been shown that the limited system in question de-
termines B exclusively.

Thus, neither of the cases thus far cited can be said to furnish a
real case of subjectivity. We require an instance of something
which is a part of consciousness, which implies consciousness, or
is exclusively determined thereby.

3. The presence of some elements alone in one complex is dependent
on the selective action of consciousness. — There is, in other words, a
privative character attaching to the assemblage of contents of
consciousness which can be accounted for in no other way but by
the sensibilities, threshold, attention, etc., of a sentient organism,
or by the organism's selective interest.

The limited manifold of a mind's contents appears upon retro-
spection, when it is contrasted with the larger manifold from which
it is taken.1 One then learns to distinguish what-one-was-conscious-
of, from the complete environment of consciousness. But intro-
spection alone does not reveal the causes of selection, the conditions
making the difference between what does and what does not 'get
into' consciousness. The difference is evidently not one of ele-
mentary constituents ; for the contents of consciousness are inter-
changeable with the contents of the surrounding field. There is
nothing in the inherent nature or quality of 'greenness,' 'round-
ness,' 'relation,' etc., that determines either their presence or their
absence from among the contents of a mind. But an explanation
is found in the capacities and action of the organism. Thus in

1 In an article entitled Conceptions and Misconceptions of Consciousness,
Psychological Rev., 1904, 11, 282-296, I attempted to justify and illustrate this
assertion. James (Essays in Radical Empiricism, 24) and Lovejoy (Reflections
of a Temporalist on the New Realism, J. of Phil., Psychol., etc., 1911, 8, 594)
have expressed their agreement with this view. But in the article mentioned I
did not sufficiently account for the limited manifold.

DEPENDENCE ON A PRIMARY CONSCIOUSNESS 139

order that 'greenness' shall appear among the contents of mind M,
M must possess the capacity of vision developed to the point of
color discrimination ; M must be attending to ' greenness ' ; and
greenness must be relevant to some interest which is moving M.
These and other like conditions determine the difference between
greenness 'in,' and greenness or any other element 'out,' of con-
sciousness; or between the class of the 'ins' and the class of the
'outs.'

Thus the complex of which one is conscious is determined as a
limited or partial complex by the relationship which it sustains
to the subject of consciousness. Much that is supposed to be
dependent on consciousness in a more drastic sense can be explained
by this principle, notably the cases of spatial perspective and tem-
perature relativity. These are selections from the full geometrical
or thermal field, and are not created, but only picked out by the
position or state of the sentient organism.1

4. The presence of some elements together in one complex is depend-
ent on the combining action of consciousness. — The same condi-
tions which determine the inclusion of contents of mind, determine
also their partnership. The physical, physiological, biological, and
ethical principles which determine the entrance of 'greenness' into
the complex of M's content, determine also what shall be there
with 'greenness' ; for example, the complex 'greenness' and 'round-
ness,' as co-contents.

This may be a matter of little or of great importance for the
elements so correlated. In other words, it may mean no more than
the bare fact of fellow-membership, the peculiar cross-relation
among contents ; or it may mediate some of the further varieties
of dependence enumerated below. Thus the relation sustained by
A and B within the content-field may also be wholly determined by
other causes ; as when, e.g., A and B are perceived in their natural
spatial relations. Or, A and B may be combined in a new way;
that is, in a way determined by the agency of consciousness ex-

1 Cf. Holt, below, 303 et al.; and Pitkin, below, 393.

140 A REALISTIC THEORY OF INDEPENDENCE

clusively. It is this which occurs when the imagination is said
to be ' creative.' Or A and B, through their co-presence as content,
may indirectly acquire new relations, such as 'meaning' ; and these
new relations may be of crucial importance for A and B. They
may, for example, bring about the employment of A and B in the
subsequent operations of M . In short, as fellow-members of one
consciousness, A and B may begin a new epoch in their careers
through being brought within the play of practical and social
forces.

Thus far, then, we have recognized two instances of genuine
dependence on consciousness ; both of which can be characterized
as content-patterns. Content-complexes possess an individuality
both as respects what they include, and as respects what they in-
clude together; and in both respects are determined exclusively
by the agency (selective and combining) of the subject of conscious-
ness.

5. Value is dependent on consciousness. — This is a matter on
which neo-realists are by no means agreed.1 To the present writer,
however, it seems evident that value is a function of desire. This
does not mean that the 'precious metal,' gold, is dependent on
desire; but only that its preciousness is thus dependent. It de-
rives its economic value from the ' demand ' for it ; 2 and its decora-
tive value from the sensuous gratification which it affords. Gold
in other respects may be, and is, for the most part, entirely inde-
pendent of consciousness. Gold can, on this ground, be declared
to be dependent on consciousness, only provided its preciousness
is included in its definition. There is a physical or chemical gold
that is not precious ; which is, in other words, definable and de-
terminable without reference to the part it plays in economic and
aesthetic life.

1 Moore and Russell, e.g., hold that 'good* is independent of consciousness. Cf.
Moore, G. E., Principia Ethica, 137. For my own view, cf. The Moral Economy,
Ch. I, and Present Philosophical Tendencies, Ch. XIV.

1 Mere rarity does not, of course, give value to a thing unless there is a demand
for it.

DEPENDENCE ON A PRIMARY CONSCIOUSNESS 141

Whether value is dependent on knowledge, or not, is another
question. Though contrary to my belief, I am perfectly willing,
for the purposes of the present argument, to concede that a thing
cannot be desired without being known. In that case, then, value
does depend on knowledge of the object possessing value. But
in any case, as we shall see presently, value does not depend on
being known to be such. That is, it is possible to desire without
knowing that one desires. And it is the primary relation of desire
that endows a thing with value, whether the relationship itself be
known or not. The important fact is that A's value is A's being
desired by M, N, or some entity capable of desire. If one then
assumes that desire is a variety of consciousness, and that an entity
capable of desire is a subject of consciousness, it follows that A's
value is dependent on consciousness ; not only on the primary de-
siring act which directly endows it with value, but on whatever
other conditions in the conscious subject, such as the presence of
other desires, affect that primary desire.

6. Works of art are dependent on consciousness. — By 'work of
art ' I mean whatever complex is caused by the physical organism
acting in pursuance of its interests. Such a complex may owe its
internal and external arrangement to the organism's action, to a
greater or lesser extent. It may be simply used 'as it is' ; or it may
be moved, divided, redistributed, or brought into new physical
configurations. In any case, what happens to it exclusively in
consequence of the purposive action of life, is dependent on con-
sciousness. Consciousness is the means of bringing things within
the range of purposive action. It determines the limits of the en-
vironment 'taken account of/ as distinguished from the total en-
vironment. The range of consciousness defines a field of things
liable to action. A thing 'noticed' is a thing that can be avoided,
used, or remade, as exigencies may require. Furthermore, the
actual dealings of the organism with such objects, the process of
art itself, is guided by consciousness.

It is important to note that the modification of the object is not
in this case directly due to consciousness. By consciousness the

142 A REALISTIC THEORY OF INDEPENDENCE

object is brought within reach of other causal agencies. It is made
dependent on the body by the action of the mind. Again, it does
not follow that any object A which is subsequently modified by the
action of the body is unqualifiedly dependent on the mind. It is
dependent thereon only for what is 'done' with it or 'made' with
it. It is quite open to us to suppose that A is independent of its
being manipulated at all. Furthermore, if for any reason the modi-
fying action of the organism were not to take place, A would never
enter upon the phase of dependence. Consciousness renders its
objects dependent in this sense only in so far as it is followed up
by some physical operation involving the objects. In short, works
of art are dependent on consciousness for, and only for, whatever
there is of art in them.

7. Higher complexes, such as history, society, life, or reflective
thought, are dependent on consciousness. — Whatever complex con-
tains consciousness as one of its components, or whatever set of
premises implies consciousness as its necessary conclusion, is evi-
dently dependent on it; and I have cited the most obvious in-
stances that occur to me. It is, of course, to be remembered that
while these complexes as wholes or sets of premises depend on
consciousness, it does not follow that the several components
depend on consciousness. Thus while a municipality depends on
consciousness, because it contains or implies it, its bricks and
mortar do not share this dependence. If they are dependent, it
must be for some other reason, such, e.g., as their being works of art.

'Reflective thought,' as defined by the pragmatists, affords,
perhaps, the most instructive instance. 'Reflective thought' is a
complex process in which one 'bit of experience' means, or is
'idea-of another. In order that A shall mean B, A, at least,
must be 'experienced.' Whether we express this in terms of the
whole-part relation or in terms of implication, in any case the
meaning process depends on the simpler process of experiencing.
It does not, however, follow that experiencing is dependent on
meaning, or that the thing experienced is dependent on being
either experience or idea.

DEPENDENCE ON A PRIMARY CONSCIOUSNESS 143

It will be observed that in the above discussion a realm of sub-
jectivity has been explicitly admitted and defined.1 If there
be a polemical virtue in admitting it, there is a constructive
virtue in attempting to define it. It is no part of realism, as I
understand it, to reject "purely subjective existence," out of
hand ; 2 but rather to be more specific about it, and above all to
avoid hypostasizing the facts of subjectivity into a substance, or
into a new continent inhabited only by subjectives. But perhaps
I should say that the realist does deny 'purely subjective existence' ;
for any realist would deny that there is anything that is sub-
jective through and through. A subjective complex can always be
analyzed into elements, or even into lesser complexes that are
'objective.' Subjectivity, in the sense of exclusive determination
by a subject of consciousness, attaches only to certain relation-
ships or complexes in their solidarity.

Furthermore, realism does not consist essentially either in the
denial or the assertion of subjectivity; but in the assertion that
there are cases of entities, simple and complex, that are inde-
pendent of consciousness. ' Epistemological monism' means that
when perceived, things are directly and identically present in
consciousness ; in virtue of being perceived, they constitute what
is called content. And realism adds the further assertion that,
in certain notable cases, at least, things are none the less inde-
pendent for being so perceived. Thus the case for realism rests
on showing that to be content of a mind, is not to be dependent
on a mind.

The questions of error, illusion, hallucination, dreams, etc., all
raise new issues; and these issues are dealt with in other parts
of the present volume. Subjectivity is not error. The whole
point of error lies in the difference between building 'air-castles,'
and mistaking them for something more substantial. But in the

1 This list of cases of subjectivity is not intended to be complete, but only illus-
trative. It is not intended to exclude, e.g., the possibility of cases in which the
body perceived is physically modified by the sentient organism in the act of per-
ceiving it. Cf. above, 135. * Lovejoy, op. cit., 597.

144 A REALISTIC THEORY OF INDEPENDENCE

concluding portion of the present chapter I shall present cer-
tain considerations that are at least relevant to the question of
error. For I propose to point out that subjectivity itself possesses
a certain independence as respects a secondary consciousness, a
superadded knowledge of it ; which will at least show that there is
a radical difference between bare subjectivity itself, and that
vicious subjectivity through which the term is confused with the
misfortunes of cognition.

VII

THE INDEPENDENCE OF SUBJECTIVITY ON A SECONDARY
CONSCIOUSNESS

1. The subject of consciousness is independent of being known. —
The subject of consciousness is of course a case of subjectivity.
I do not mean that what assumes the role of subject is dependent
on that role, but that it is dependent in that r61e. It is only in
a loose sense that one can speak of dependence here at all ; it is,
more strictly, a case of identity, and I have not thought it worth
while to cite this as a case of dependence. The important fact is
that a subject need not be known. A subject may be the condition
of the content-status assumed by its content without itself assum-
ing such a status.

Thus there may be consciousness without self-consciousness.
Idealists have always accorded a partial assent to this conclusion,
in that they have denied the subject a place among its own con-
tent. They have called attention to the fact that the content-
manifold as a whole is the passive correlate of an active subject.
But they have felt called on to provide some unique way in which
the subject may be known without becoming content. The
result has been to formulate a contradiction that they have never
succeeded in relieving. For to refer to the subject as known,
whatever peculiarity may attach to knowledge in this instance, is
evidently to put it on the passive side of the correlation. The

THE INDEPENDENCE OF SUBJECTIVITY 145

difficulty is as gratuitous as it is insuperable. There is no reason
whatsoever for supposing that whatever knows, must be known.
There is no reason why the subject, in order to condition a content-
manifold, should itself lie within the manifold.

On the other hand, there is no reason why the subject of the
cognitive relationship M-A should not be known, through sustain-
ing a like relation to another subject N. In other words, there is
no difficulty in supposing a complex relationship, N-M-A, where
A is content of M , and M of N. Thus it is entirely consistent
that the acting organism which does not appear within the field
of its own objects should nevertheless appear within the field of
objects of a second organism.

2. One consciousness may be independent of another. — Two in-
dividual units of consciousness may be dependent in that one in-
cludes or implies the other, or in that the two are mutually and ex-
clusively determined. But on the same grounds one is independent
of the other in so far as it does not include or imply it, or sustain
relations of exclusive determination with it. And such cases
abound.

Suppose, for example, as is universally the case when the ques-
tion of solipsism is under discussion, each of two consciousnesses
lays claim to the other. Thus M finds N as a part of itself, and N,
M. Then in so far as M includes N as its object, it does not follow,
as is commonly argued, that N depends on M , but 'rather that M
depends on N. Argued on these grounds the conclusion would be
just the reverse of solipsism and would result in an utter self-
abnegation. As a matter of fact, however, it is possible in the
majority of such cases to define a narrower M which excludes N,
and is independent of it. M does not depend on N unless M is de-
fined to exclude everything which happens to it. M may, however,
be a soul-substance, or a constant nucleus of states, or a central
purpose, and is then not dependent on N unless these include N or
are defined and determined by N exclusively. And the same holds
true of N in its relations with M. The root of the solipsistic ab-
surdity is a failure to remember that such a situation is mutual.

L

146 A REALISTIC THEORY OF INDEPENDENCE

If there is ground for asserting one witness, there is the same ground
for asserting a 'cloud of witnesses.' And if any individual con-
sciousness can set up independently and then proceed to annex
other consciousnesses, then these other consciousnesses enjoy the
same right. If the solipsistic train of reasoning is generalized, it
destroys itself. But as a dialectical argument it is always general-
ized. It postulates an independence in the case of one conscious-
ness which cannot on principle be denied to others, and which is
self-contradictory if generalized.

Philosophies which emphasize the unique certainty and inde-
pendence of the self make much of the argument from analogy.
We are said to know ourselves first, and then infer the existence
of other selves from the similarity of their behavior to our own.
But as Mr. Moore has very effectively pointed out, the argument
from analogy in this case assumes the independence of physical
facts.1 Another's consciousness is said to be inferred from his be-
havior. But such an inferred consciousness can be regarded as
another's only provided I regard his behavior as a physical reality.
If I construe another's behavior as my perception, then I can
infer only my own consciousness, and not his. The analogy from
which I argue may be formulated in either one of two ways. I
may say that when I have a perception of a bodily contortion, I
have also a feeling of my own pain. Or I may say that when there
is a certain bodily contortion there is correlated with that body a
feeling of pain. And the argument from analogy leads to different
conclusions in the two cases. In the first case having had a second
and similar perception of bodily contortion, I expect myself to have
a similar feeling of pain. I argue, in other words, from a relation
between my perceptions and my feelings. But in the second case,
observing a second and similar bodily contortion, I infer a similar
feeling of pain to be correlated with that body, as mine is corre-
lated with my body. In the one case I argue from my perceptions
to my feelings, and never get outside the operation of the laws of

1 Moore, G. E., The Nature and Reality of Objects of Perception, Proceedings of
the Aristotelian Society, N. S., 1906. 6, 111-121'.

THE INDEPENDENCE OF SUBJECTIVITY 147

my own consciousness ; in the other case, I argue from bodies to
feelings, and have from the beginning regarded my own conscious-
ness as only an instance of such correlation. As Mr. Moore justly
concludes, the existence of other consciousness, if it is inferred at
all, is inferred from bodies as such, and not from the states of the
mind which performs the inference.

In my own view, other consciousness is not inferred at all. It
is observed precisely as physical phenomena are observed. It
consists in a complex relation between a sentient and interested
organism and some parts of its environment ; and its independence
of another onlooking self is only a special case of the independence
of physical events on the observation of them.

In any case, there is no reason for doubting the mutual independ-
ence of selves as respects their mere consciousness of one another.
They may enter into relations of whole and part or into physical
or social relations of causality; but no universality or necessity
attaches to such dependence. The only general arguments for the
dependence of one consciousness on its apprehension by another,
namely, the dialectics of solipsism, and the alleged inference of
others by analogy with self, virtually presuppose independence at
the outset.

3. Mental content is independent of introspection. — Any doubts
as to the truth of this assertion must cast suspicion on the validity
of the method of introspection. For introspection is supposed to
be the means of knowing what contents are in the mind. These
contents are in the mind by virtue of the selective action of some
subject of consciousness. But once there it is supposed that they
may be observed there by introspection. Just in proportion as
introspection itself introduces new elements, it is a source of con-
fusion. Such new factors must if possible be identified and elim-
inated. Introspection yields genuine psychological results only
in so far as it reveals that which was determined to be in mind apart
from the act of introspection itself. Otherwise it is not knowl-
edge of psychical data, but simply a psychical disturbance, which
itself constitutes a new psychical datum.

148 A REALISTIC THEORY OF INDEPENDENCE

It is very easy to fall into confusion here. It is said that psychi-
cal facts appear only on retrospection when they are contrasted
with the non-psychical or objective. "The instant field of the
present," says James, "is only virtually or potentially either object
or subject as yet." The 'state of mind' is "first treated explicitly
as such in retrospection." 1 But to be state of mind, and to be
"treated as such," are very different matters. To say that the
limited field of ' states ' is first discovered when a mind doubles on
itself, and sees around its former limits, is one thing, and is substan-
tially correct, so far as any single mind is concerned. But to say
that the limits did not exist until they were 'seen around,' is a very
different and entirely unjustifiable assertion. The " instant field of
the present" is a potential object of introspection, but only because
it already possesses a psychical character. Did I feel at liberty
to employ a conception of mind which I have not had an opportu-
nity of expounding or justifying in the present article,2 1 should go
further, and say that even the knowledge of mental contents is
independent of introspection, in that mental content may be di-
rectly, and in some instances more accurately, observed by a
second mind.

4. Value is independent of judgments about value. — Value, as
we have seen, consists in a relation to desire. In order that a thing
shall be valuable it must be object-of -desire. But it does not
follow that the complex relationship, object-of-desire, must itself
be object of consciousness. For the purposes of ethical inquiry,
needs, desires, demands, etc., together with their objects, must be
regarded as facts to be thought about, and if possible, explained
systematically. But they are no more dependent on ethical thought
about them, than physical events are dependent on physics.

It is only as facts or events that desires are final or infallible. If
M desires A, then he does, and there's an end of the matter. But
if N, or M himself, thinks that M desires A, then he is liable to
error. Were value to consist in thought about value, we might

1 Essays in Radical Empiricism, 23, 24.

J I refer to the view discussed in Ch. XII of my Present Philosophical Tendencies.

THE INDEPENDENCE OF SUBJECTIVITY 149

be driven by the fear of a skeptical relativism into the haven of
an absolute thinker that should standardize values. To say that
the good is what anybody thinks good, is both dialectically and em-
pirically untenable. But to say that the good is what anybody
desires, is simply to offer a definition which is both in agreement
with fact and logically innocent. There is no more dialectical
difficulty in this than in saying that an instrument is what anybody
uses, or a footing what anybody stands on. The current confusion
of what should be a perfectly evident truth, is due mainly to the
invention and wide use of terms like 'evaluation,' 'appreciation,'
'affective judgment,' etc., in which the notions of desire and of
judgment are fused together into a vicious equivocation.

5. Perception and simple apprehension are independent of
reflective thought. — Reflective thought in which A means or is
idea of B, requires that one or both shall be 'experienced'; and
therefore contains or implies whatever relation is in question when
it is said that A or B is 'experienced.' The relation of meaning or
ideation is a relation within a manifold of elements, some of which
already belong to consciousness in a more primitive sense. The
reverse dependence, however, does not hold. In order that con-
cept A shall be apprehended, it does not require to be used as an
idea; in order that body B shall be, or be perceived, it does not
require to be meant by an idea.

There is good ground, therefore, for the pragmatist's polemic
against identifying things with their thought-status, or intellectual
form. Things have an independent footing in an immediate or
presentative knowledge, which not only exceeds but also underlies
mediate or representative knowledge. The only fault in the
pragmatist's view is that it does not clearly and expressly take the
next step, and say as any thorough realist will say that things are
likewise independent of experience.

This consideration affords just ground for suspecting that such
writers as Dewey, for example, are not thorough-going realists.
They emphasize independence of a certain elaborately complex in-
stance of experience, the instance, namely, of discursive or mediate

150 A REALISTIC THEORY OF INDEPENDENCE

knowledge ; but they by no means make it clear that they do not
take the realm of experience itself to be all inclusive.1 And such
a generalization would certainly be contrary to realism. Further-
more, Professor Dewey's restriction of the term 'knowledge* to
the discursive process, allows most instances of selective conscious-
ness, of experiencing, to fall outside the application of his principle
of independence. Thus sensing, for example, is not knowledge;
and therefore the principle that things are independent of knowl-
edge, does not hold of the case of sensing. It would be more proper,
I think, to regard sensing as a case of knowing. Even 'smelling,'
which Professor Dewey thinks to be clearly beyond the pale, should
be so regarded; for it is evidently different from "gnawing or
poking" in that it introduces a specific content into the mind, and
so makes the mind aware of a characteristic of its environment.2
In any case it is clear that if things are universally dependent on
such experiencing as sensing, and if knowledge takes place only
within the field of experiencing, then the independence of things on
knowledge still leaves them dependent on action, or on life, or on
some such principle, which for a thorough-going realism must be
regarded as all one with knowledge.

Philosophy is here again the victim of an equivocal term. For
'experience' may be taken to mean the things experienced, or the
experience-relationship itself. To contend that experience is in-
dependent of discursive thought may still leave one well within the
ramparts of idealism. For the independent thus defined, may be
construed as the complex process of experiencing. One does not
become a realistic outlaw until one has either expressly interpreted
experience in the first sense, as things, simply, or has expressly as-
serted things to be independent of experience, in the sense of ex-
periencing or being experienced.

!Cf. Dewey, J., Reality as Experience, J. of Phil., Psychol, etc., 1906, 3,
253-257. I have discussed this matter more fully in my Present Philosophical
Tendencies, pp. 224-225, 314-316.

»Cf. Dewey, J., Brief Studies in Realism, I,J. of Phil., Psychol., etc., 1911, 8,
396 (note).

CONCLUSION 151

VIII

CONCLUSION

Our conclusion, briefly summarized, is as follows : —

1. Independence is non-dependence.

2. Dependence is not the same as relation, but is a special type
of relationship, in which the dependent contains, implies, or is ex-
clusively caused or implied by that on which it is dependent.

3. The independent may be related or not, provided it is not
related as above (2).

4. The object of consciousness is related to consciousness, but
it does not follow that it is dependent on consciousness.

5. There are entities, embracing all simples and some complexes,
which are not dependent on consciousness, because not related to
it as above (2).

6. This in no way prevents their being otherwise related to con-
sciousness.

7. There are cases of subjectivity, that is, of complexes, that as
such are dependent on consciousness.

8. Subjective complexes both contain entities that are hide-
pendent of them, and also are independent of secondary conscious
relationships into which they may enter.

A DEFENSE OF ANALYSIS

A DEFENSE OF ANALYSIS

BY EDWARD GLEASON SPAULDING

I

INTRODUCTORY

IN this essay I shall attempt both a general and a specific de-
fense of analysis. A general defense might not be exclusively
dependent upon a general realistic position; but there are re-
vealed in a defense of analysis as such many reasons which make
for the correctness of the realistic position. My specific purpose
becomes, then, not simply to defend analysis qua analysis, but
also by this means to defend the general realistic interpretation
of both whole and part. I shall, then, defend analysis as a method
of knowing which discovers entities or parts which are real in quite
the same sense as are the wholes which are analyzed. This posi-
tion may be called Analytical Realism.1

1. The Types of Analysis and of Wholes. — There are two
general types of analysis: (1) formal, and (2) experimental
or 'material/ both of which imply a relation of great im-
portance to science and philosophy, namely, the whole-part
relation.2 That which is analyzed is a whole. Analysis is
the discovery, or, possibly, the invention of parts — the parts of
the whole analyzed. Which of these analysis is — discovery or
invention, — revelation of fact or falsification — , is in reality the
central question at issue. On this point parties divide. But all
agree that that which is analyzed is in some sense a whole, and that
that to which analysis leads is in some sense a part.

1 The general evidence and proof for the realistic position are presented in the
Introduction to this volume.

1 See Perry, this volume, 107, and Russell, B. Principles of Mathematics I, 360
el passim.

155

156 DEFENSE OF ANALYSIS

By experimental analysis is meant that kind of analysis which
is made in the case, for example, of chemical compounds. These
are sometimes physically taken apart, and their constituents are
perceived or revealed in quite the same way as are they themselves,
as wholes. It is this kind of analysis that is made, to a cer-
tain extent, at least, of those entities which are dealt with in the
chemical, the physical, the biological, and, perhaps also, the
psychological laboratories. It may not be justifiable to call all
such analysis 'material,' except by analogy, but it can be called
experimental. -j

Of the same character is that analysis which accompanies our
non-scientific, perceptual, and certain conceptual processes.
Perception is itself analytical, discriminatory, selective. In it,
preceding all scientific hypothesis, there is analysis like that to
which experimentation itself leads.

By formal analysis is meant that kind which is made of such
typical wholes as the motion of a projectile, the flow of an electri-
cal current, the number-continuum, the continuity of time, etc.,
where the parts are distinguished and discovered, but nevertheless
left in situ. Some of these wholes are physically observable en-
tities and experimentally analyzable. But not exclusively so.
At a certain point other methods must be brought in, which are
identical with mathematical analysis, methods which are based
on rational principles, methods which lead to analytical results
that in many cases are directly confirmed by further experimentally
directed observation and measurement. Such analysis may be
called formal.

Examples of specific complexes which are analyzed might be
cited in great number, and so, of course, might also the specific
analyses and the results to which they lead. But, just as the types
of analysis can be distinguished, so also can a classification be
made of the complexes or wholes which are analyzable, although
to do this itself involves analysis, and constitutes a statement of
results rather than a method of proof. But the very type of analy-
sis which is involved in discriminating not only different lands of

TYPES OF ANALYSIS AND OF WHOLES 157

analysis, but different kinds of wholes, is subsequently considered
in detail in this essay, and receives its full justification.1 There
are:

1. Aggregates or collections of any number of objects in any
order, hi numerical conjunction. Thus there is the collection of
objects with which I am now concerned, namely, this chair, and
this table, and this pen, and my thoughts, and the concept 'whole,'
and 1 and 2, etc.

2. Classes formed or composed of parts which are not classes,
but which may be either organic wholes, or individuals, or simples,
or collections. Thus the atoms of carbon, all electrons, the even
integers, the rational fractions, are such wholes.

3. Classes formed or composed of subordinate classes; ex-
amples: element, number, integer, etc., which are subdivided
respectively into the classes, monovalent and bivalent element,
cardinal and ordinal number, odd and even integer.

4. Unities or organic wholes ; examples : any specific individual
chemical compound existing at some particular place and time,
any one organism, any one individual molecule or atom.

2. What is Analysis f — The wholes which are analyzed differ as
the above classification indicates, and analysis is itself either mate-
rial or formal. But what are the characteristics of the genus analysis
which, for example, is presupposed by the classification of analyses
and by the discovery that there are wholes or complexes composed of
parts ? The answer is that analysis is, doubtless, itself somewh at com-
plex and devious, and that perhaps anything more than a working
definition of it is most difficult. Analysis may be a process, but
if it is, it would appear to be such a complex one that a simple
definition of it is impossible, and that its own character can be re-
vealed only by an elaborate analysis. But an exact and precise
logical definition may not be necessary. Every one understands
in a general way what analysis is, what it means, and what it does.
To this general understanding appeal may be made, and, relying

1 See Section IV of this essay.

158 DEFENSE OF ANALYSIS

on it, it may be said that to analyze means to discover that an
entity is in some sense formed or composed of parts. Thus, as
previously asserted, analysis involves the whole-part relation.
The entity which is formed or composed of parts is for that reason
called a whole or complex, although, in some cases, its parts may
in turn be wholes or complexes, as we have seen. The position,
however, that analysis means in every, or even in any case, dis-
covery of parts, — implying that these already exist or subsist
independently of the analysis and discovery — is one which is dis-
puted. As opposed to it the position is taken by some, that to
analyze means to invent, to construct for purely practical purposes
of one kind or another, such as prediction and the control of nature,
an artificial division into parts.1 This view fits in with, or is a part
of that aspect of pragmatism which is well called the instrumental
or biological view of knowledge.2 According to it, analysis is
simply an intellectual instrument, a mode of adaptation, where
anything which works, in the sense that it does adapt the organism
to its environment, is to be accepted at least at its face value,
namely, that it does work. Only, of course, pragmatism goes
beyond this, and identifies this working with truth. Secondly,
this pragmatic view is sometimes extended to mean that reality
in general, or any specific part of reality which may be selected
and distinguished from other parts, is plastic and lends itself to
almost any kind of analysis and moulding, but that the analyses
which are or have been actually made are necessarily constructed
from a human point of view. This is the humanistic interpretation
of analysis, as well as of its counterpart, synthesis, and of philo-
sophical and scientific method in general.3 Thirdly, both of these
interpretations are compatible with that interpretation of analysis
which makes its validity identical with its verification, giving
psychological pragmatism,4 — or dependent upon and tested by

1 Bergson, H. (Mitchell, A., trans.) Creative Evolution. 1910.
1 Cf. Montague, W. P., May a Realist be a Pragmatist, J. of Phil, Paychol, etc.,
1909, 6, 460 and 485.

3 Cf. Montague, ibid., 561. « Cf. Montague, ibid., 543.

PRAGMATISM 159

its utility or value, giving logical pragmatism.1 These four posi-
tions as to the nature of analysis are open to various criticisms,
among them being those which can be made from a realistic stand-
point. For the realist regards analysis, in the great majority of
cases at least, to be the discovery in a whole of elements or parts
which exist or subsist independently of the analysis and discovery.
He grants, of course, that analysis is useful, admits, perhaps, that
there are few things more so, but he finds that such usefulness
presupposes the realistic interpretation of analysis.2 So, also,
he admits that the validity of analysis may, to a certain ex-
tent at least, be tested by its value, although so to do involves
many difficulties and ambiguities as to what standard of value
shall be selected. Further, those aspects of wholes which
shall be discovered as parts may be dependent for their discovery
upon distinctly human interests. For various and sundry reasons,
namely, for those of scientific tradition, of religious feeling, of aes-
thetic appreciation, — all human ' things ' — the actual analyses
which men are interested in making may be only some out of many
that are possible; they may be selective. However, the realist
can make all of these admissions and yet retain his realism. In-
deed, not only this, but he holds, further, that he can show that
all of these positions which he is willing to admit demand his
realism rather than controvert it. It is only to the position,
that the validity of analysis is in any specific case identical with
its verification, that the realist is fundamentally opposed. For
such a position means, he holds, subjective idealism.3 Of that he
can and will have none.

But the pragmatic interpretation is not the only one. There
is at least one very distinct type of attack, which, at the same time
that it is pragmatic to a certain extent, goes quite beyond this
interpretation. It is an attack which, while in direct opposition to
the realistic view of the results of analysis as discovery, is, never-

1 Cf. Montague, ibid., 561. * Montague, in the four papers just cited.

8 Montague, ibid., 543.

160 DEFENSE OF ANALYSIS

theless, realistic in its interpretation of the whole which is analyzed.
It opposes one kind of realism only to reinstate another, using
rationalism as its method of attack, and mysticism in support of
the realism which it does accept. Such an attack on analysis
has been recently advanced by Bergson,1 and has attracted much
attention. A similar attack was made by Bradley 2 less recently.

The very core of this attack is the claim that analysis is identical
with falsification in that very precise sense in which this term (fal-
sification) can be defined, — namely, as involving contradiction.
According to this attack, analysis is the finding, or inventing, or
constructing, or discovering — or what-not — (in a whole) of
parts which in a certain one, or in a few, or in many respects are
the contradictories of the whole. For example, it holds that the
analysis of motion leads to 'rests,' as terms,3 but that, since rest
is the contradictory of motion, either the analysis or the intuition
of the whole, of the motion qua motion, is false, and the former
alternative is chosen. For Bergson, such contradictory parts
are invented -1- by intellectual processes, and, were he consistent,
he would have to admit that anything short of One all-inclusive,
interpenetrated, Evolving-whole is self-contradictory, and so false
and not real. For Bradley, the parts are produced, or even pos-
sibly discovered, by intellectual processes, but, since they in-
volve contradiction, anything short of One Absolute is false and
unreal. Briefly, the recipe for both of these philosophers seems
to be, Self -contradictory 'things' are what we find or get when we
reason; therefore, to get at truth, at reality, avoid reason, and
use feeling, intuition, ecstasy, absorption !

It is not my purpose to present the history either of the attack
or of the defense. But the point must be made, that all the at-
tacks on analysis are made by methods which themselves involve
analysis or are analytical. The analysis which is attacked may
be different from the analysis which is used in attacking, and so,

1 Creative Evolution, and Time and Free Will (Pogson, F. L., trans.) 1910.

* Bradley, F. H., Appearance and Reality, 1894.

* Bergson. Creative Evolution, 163.

CHARACTER OF ANALYSIS 161

conceivably, a specific attack might be valid. But if it were, it
could serve to invalidate only certain kinds of analysis, to show
their limitations, etc., but the invalidation could not be universal.
The validity of at least the means of attack would, willy-nilly, be
presupposed. Some analysis, at some point, would be exempt
from successful attack and criticism. Then the supposition would
remain, that, if some analysis is valid, all might be, and that at
best only certain specific analyses could be in error, but that analy-
sis qua analysis could not be invalidated.1

But to return to our question as to what analysis is. Given a
whole which, for one reason or another, is known to be analyzable,
then analysis reveals parts, but it also reveals the relations which
relate and so organize these parts into some kind of whole. Con-
sider also those properties which, in some cases, the whole, as a
whole, may have different from those of the parts. Of course,
analysis reveals these also. The analysis may be incomplete in the
sense that there may be further parts, that is, parts of parts, which
are not yet revealed ; but, if the analysis is incomplete only in this
sense, that is, if there have been revealed parts, their organizing rela-
tions, and, in some cases, the possibly' specific properties of the
whole, then the analysis may be said to be adequate. It exhausts
the whole up to the point that it reaches, in that, while the speci-
fication of all that the analysis reveals does not specify the whole,
the whole, nevertheless, is the parts and their properties and the
relations relating the parts and the possibly specific properties of
the whole. There may be further parts of parts, more properties,
more relations to be revealed, but this of itself does not invalidate
the position that the properties of the parts and the generating
relations which are revealed are quite as real as is the whole which
is analyzed, are not the contradictory of the whole, and exist, or
subsist, independently of the discovery and of the specification.

This, then, is one meaning of analysis, a more precise and ade-
quate one than the one previously given. It is, however, another

1 Cf. my paper, The Postulates of a Self-critical Epistemology, Phil. Review, 18,
615.

It

162 DEFENSE OF ANALYSIS

definition which is tacitly made by the attacking party of the
Bergsonian, and, sometimes, of the Bradleian type, although not
by the pragmatist, the definition, namely, by which analysis is
held to give only terms and in which no cognizance is taken of the
organizing relations. Advantage is taken of this omission to con-
trast the terms and their properties with the whole and its prop-
erties, and thus, through this neglect, to find the looked-for
contradiction between part and whole, and so the falsifying
character of the analysis.

II

COLLECTIONS AND ENUMERATIVE ANALYSIS

THE first type of whole which can be distinguished is that
whole which is simply an aggregate, or collection. Among its
parts there may be similarities and differences of various degrees.
So, too, may the types of relations which exist, or subsist, be-
tween these parts be many or few, similar or widely different.
Certain specific relations may exist, or subsist, between certain
parts, others between other parts, but in any case, whatever be
the parts, and whatever be the relations, the parts form a col-
lection or aggregate in that all the parts are related to one another
by the relation which is expressed by and.1 The process of enu-
meration can be started with some parts, whether or not it is com-
pleted or completable. In this sense, all the parts of any col-
lection, or any number of parts of any collection, form a whole ;
they are denumerable, and so stand in one to one correlation with
the cardinal integers. This relation to the members of at least
one other class, the cardinal integers, the parts of a mere collection
have, whether or not between all of them as parts any one other
relation holds exclusively or not.

In this connection I may quote Mr. Bertrand Russell. "When

1 1 shall regard and as expressing a relation, namely, the numerical conjunctive
relation, although there are departures from this usage. Cf. Russell, Principles
of Mathematics, 71.

COLLECTIONS AND ENUMERATIVE ANALYSIS 163

a class is regarded as defined by the enumeration of its terms, it is
called a collection. By a collection I mean what is conveyed by
'A andB' or 'A and B and C,' or any other enumeration of definite
terms. The collection is defined by the actual mention of the
terms, and the terms are connected by and. It would seem that
and represents a fundamental way of combining terms, and that
just this way of combination is essential if anything is to result of
which a number other than 1 can be asserted." J "Every pair of
terms, without exception, can be combined in the manner indi-
cated by 'A and B,' and if neither A nor B be many, then A and
B are two. A and B may be any conceivable entities, any possible
object of thought, they may be points or numbers or true or false
propositions or events or people, in short anything that can be
counted. A teaspoon and the number 3, or a chimsera and a four-
dimensional space, are certainly two. Thus no restriction what-
ever is to be placed on A and B. It should be observed that A
and B need not exist, but must, like anything that can be men-
tioned, have Being. The distinction of Being and existence is
important, and is well illustrated by the process of counting.
What can be counted must be something, and must certainly be,
though it need by no means be possessed of the further privilege of
existence. Thus what we demand of the terms of our collection is
merely that each should be an entity. " 2

In this sense, then, anything, taken with at least one other 'some-
thing/ and these two with another something, and so on, form a whole.
Accordingly, there is an aggregate or collection of all entities — of
all terms, and relations, and classes, and concepts, and propositions,
etc., which exist, or subsist. Such an aggregate may be called the
universe, and is, in just the sense defined, a whole. It has parts,
and its parts are connected by numerical conjunction, but that
does not preclude the existence or subsistence both of wide dif-
ferences and of great similarities among its parts, and, in the latter
case, of classes whose parts are related by one or more specific
relations; nor does it preclude the independence both of terms

1 ibid., 69. » tbid., 71.

164 DEFENSE OF ANALYSIS

from one another and of classes, whether closely similar or widely
different from one another, at the same time that these terms and
classes are related. Relatedness and independence are quite com-
patible,1 and that totality of entities which is the universe is quite as
consistent with pluralism as with monism, with independence and
great differences as with dependence and one fundamental simi-
larity.

In the case, then, of the first type of whole, the terms are simply
the parts which are enumerated, and the chief organizing relation
is simply that which is expressed by the connective and. Any whole
which is an aggregate may also be more than this, and, if it is,
there are other organizing relations than and. But a whole may
be a mere aggregate notwithstanding certain other relations, such as
those of similarity and difference, among its parts. Such a whole,
however, does not seem on the basis of genuinely empirical investiga-
tion to possess any distinctive properties as a whole ; it is simply its
parts, with their properties, summed and numerically conjoined.
Nevertheless, on some other basis, that of some assumed, specula-
tive principle, from whose standpoint the attack is made, the col-
lective whole, especially that collective whole which is called 'the
universe/ is made to have properties which its parts do not have,
and to be of such a character as a whole that it cannot be analyzed
without falsification.

One has only to consider the attacks of Bergson and Bradley to
find this statement confirmed. Each of these philosophers attacks
that ordinary conceptual analysis which is crystallized in the lan-
guage of everyday discourse, and on the basis of which we can
make an enumeration of kinds. But each also attacks that con-
ceptual analysis which is identical in part with science.2

For each of these attackers of analysis, the real world, ultimate

1 See Perry, this volume; 113; also The Program and First Platform of Six
Realists, (See Appendix.) /. of Phil., Psychol., etc., 1910, 7, 393 ; also Introduction
to this volume.

1 Bergson, Creative Evolution, 9, 160-163, 303 et passim, and Bradley all
through his Appearance and Reality.

BERGSON AND BRADLEY 165

reality, cannot be disclosed by analysis — if we take their explicit
claims and ignore their tacit usage and presuppositions. It is
not a collection, not a class, rfot an organic unity, but is in some
sense just One — One Absolute or One Evolution. Analysis can
at best have only a pragmatic justification, for, false though it is,
ironically it can serve one finite part of this universe, namely, one
organism, in adapting itself to other parts, and in finding out that
they are false and that only the whole is true! Why, now, all
this subversion of the results of both common sense and science
and of much philosophy ? By reason of what are the conclusions
of these philosophers reached — for they do reason, do use prem-
ises, do analyze ?

The position, then, as to the nature of wholes, upon which this
attack on the possibility of their analysis is based, is shown by
only a slight inspection to be nothing more or less than that very
theory which the realist finds used so frequently by his opponents,
namely, the theory of 'internal relations.' The attacking party
grants that wholes are manifolds and complexes 1 — in some sense, —
but holds that the parts or elements are all constituted by their
relations to all other parts in the complex. Briefly, there is a uni-
versal interpenetration ! But this is the theory of internal re-
lations, or at least one aspect of this theory, namely, that which
may be called the 'modification' or 'constitutive' aspect of the
theory in distinction from the 'underlying-reality' aspect.2 Ac-
cording to the internal theory in general, as inclusive of both 'as-
pects,' independence 3 precludes relatedness, and conversely.
But relatedness is undeniable fact. Therefore, it is concluded,
related terms are mutually dependent. The next question is,
What is the nature of this dependence? According to one theory,
the constitutive, this mutual dependence can only mean that
each term makes a difference to the other and, therefore, consti-

1 See, e.g. Bergson, ibid., 162.

2 Cf. my papers, The Logical Structure of Self-Refuting Systems, Phil. Rev., 19,
3, 276, and 6, 610; and Russell, The Basis of Realism, /. of Phil., Psychol., etc.,
8, 158.

8 Cf. Perry, this volume, loc. cil.

166 DEFENSE OF ANALYSIS

tutes it, at least in part. The two terms in relation form a unity,
a whole, of some kind. But then, if another term is related to
this unity and to each of its terms, and so on, and so on, very clearly
each so-called term becomes or is infinitely complex. Manifestly
it becomes most difficult if, indeed, not impossible, to discover
what a term is or to identify such. Seemingly, everything is a
complex, even those ultimate terms which such a complex would
seem to imply.

According to the other aspect of the 'internal theory/ the
mutual dependence of two related terms implies a unity, but this
unity cannot exist, or subsist, as identical with the terms, for they
are two. Therefore the unity exists, or subsists, as an underlying or
transcendent entity, whose function it is to mediate the relation. By
using contradictory terms, for example, self and not-self, this unity
can be made all inclusive and yet One. However, at this point, this
second aspect of the internal theory does not concern us greatly.
That which does is the constitutive, the interpenetrative aspect,
for it is upon this that at least one part of the attack on analysis
is based. As confirming this assertion we may cite M. Bergson's
statement that "The distinct outlines which we see in an object,
and which give it its individuality are the plan of our eventual
actions that is sent back to our eyes, as though by a mirror, when
we see the surfaces and edges of things. Suppress this action,
and with it consequently those main directions which by percep-
tion are traced out for it in the entanglement of the real, and the
individuality of the body is reabsorbed in the universal inter-
action which, without doubt, is reality itself." l

Clearly, however, this position involves many difficulties. The
argument both for it and for the other 'aspect' is grounded on the
attempt to show that by the opposite view, the theory of external
relations, in accordance with which relatedness and independence
are quite compatible, the relatedness of terms cannot be explained.
But this attempt fails ! 2 In fact, the internal theory in either of

1 Creative Evolution, 11 (italics mine). Cf. also, ibid., 162, 188, 338-340.
* Cf. my two papers, just cited.

THEORIES OF RELATIONS 167

its aspects can be stated, argued, and known as a theory only at
the cost of using the theory of external relations, and so, of limiting
itself. Thus, in strict consistency with the constitutive theory,
it is impossible to find or 'pick out/ or identify any entity as a
genuine term ; but the theory is stated, argued, and known —
supposedly as an objective theory — and terms and propositions
are identified as just those terms and propositions ; and principles
of proof are accepted. To recognize and identify any so-called
term is really to tacitly assert that, even if the term is related to
all other terms, it can nevertheless be identified as just that term
without identifying all its relations, and that it is in this respect
independent of these relations. Again, the assumption that the
theory is a true, known theory presupposes that, although it is related
to the knowing, it is independent of the knowing. Further, even
granting that it does apply to some terms and relations, it requires
the theory of external relations in that it demands (1) terms —
ultimate terms — to be modified, or to constitute other terms
through relations, and (2) that these terms can be got at in some
way. The question, however, in which we are chiefly interested
is not whether the internal theory has no application, but simply
whether this application can be universal. Evidently it can
not be, for the simple reason that observation shows that it does
not fit all cases of relations. Clearly, then, it is hazardous
to argue, on the basis of its assumed universality — for all rela-
tions between all entities, — that analysis is impossible. And yet
this is exactly what is done when everything is put into a whole in
which interpenetration is made universal.

The fact is that the theory of internal relations does not have
a universal application. The infinite complexity of terms does
not exist, or subsist. Specific reasons for the incompleteness, or
the inadequacy, or even the falsifying-character of analysis, —
specific reasons other than this infinitely-interpenetrating, com-
plex constitution of terms — might conceivably be advanced in
specific cases and found to hold good. They might hold good in
some cases, and not in others ; perhaps in all, but possibly in none.

168 DEFENSE OF ANALYSIS

Which is the case, only detailed empirical investigation can show.
But to hold on the basis of the arbitrary application of the con-
stitutive theory of relations that the analysis of a whole is im-
possible, is a self-limiting, a self-contradictory position.

My defense of analysis will proceed, then, along the lines just
indicated. The attacking party does not and cannot hold consis-
tently to the constitutive aspect of the theory of internal relations ;
he cannot make this theory universal, so as to base his attack
on all analysis qua analysis on it. Nor can he consistently base
his attack on its own outcome. My defense, then, will endeavor
to show that, dismissing that argument against analysis which is
derived from the constitutive theory — as it really is dismissed
logically and involuntarily by the attacking party himself at a cer-
tain point, — the adequacy and validity of analysis can be demon-
strated if both the terms and the organizing relations, to whose
discovery analysis also leads, are considered. The attack which
is really to be met is, then, not that which is grounded on the
constitutive theory of relations, but, rather, the position that analy-
sis leads to terms which are the contradictory of the whole. This
position becomes the real argument of the attacking party. But this
attack is based on the ignoring of the organizing relations, or, in
some cases, on making a false analysis — which is, of course, easily
attacked, or — what is much the same — on a misstatement of the
actual results of correct analysis. In the latter case the remedy
is easy to apply. Consider the analysis as it is actually made, in its
correct form, and its supposed falsifying character disappears. In
the former case, consider the organizing relations, and again the fal-
sification disappears. Consider both terms and relations and the
properties of the whole which may be left over, but which are re-
vealed by analysis, and the analysis becomes adequate at the same
time that there is opportunity for that 'creative evolution/ * for
that creative synthesis which some of the attacking party emphasize
so strongly, but which is not dependent, for its acceptance, upon the
validity of their attack.

1 Cf. Bergson, op. cit., 217, el passim.

THE SECOND TYPE OF WHOLE 169

Enough has been presented to show what our conclusions must
be with reference both to the first kind or type of whole, the aggre-
gate or collection, and to the analysis with which the enumera-
tion of the parts of such wholes is identical. This analysis stands
as perfectly valid. No argument other than that one which is
formed purely speculatively and by analogy, and which is self-
refuting, can be advanced to show either its intrinsically falsify-
ing or inaccurate character. There is no strictly empirical argu-
ment which leads to these conclusions. But further, just as no
strictly empirical basis is found for applying either aspect of the
theory of internal relations to the parts of a collection, so there is
no empirical basis for applying either of these aspects to the cog-
nitive situation which is involved in the act of enumerating, and
of analyzing. Rather, the empirical evidence is the other way,
namely, that the theory of external relations applies to this situa-
tion.1 Then there is no opportunity for accepting the analysis
as valid, as analysis, and yet of interpreting it from some non-
realistic standpoint. Our conclusion is, therefore, that this first
type of analysis stands as valid qua analysis, and that also it is
to be interpreted only realistically.

Ill

THE SECOND TYPE OF WHOLE (SPACE, TIME, ETC.) AND ITS

ANALYSIS

THAT the conclusion just reached for the first type must be
accepted for the analysis of the other kinds of wholes is the posi-
tion also held by the realist. But this must be demonstrated
in detail for each kind. We proceed, then, to the considera-
tion of the second type. In wholes of this type there are parts,
revealed by analysis, all of which, in any one whole, are similar
in at least one, possibly in many respects, and are related by one or
more common relations other than that of numerical conjunction.

1 This volume, Introduction.

170 DEFENSE OF ANALYSIS

Now it is against certain specific types of the analysis of wholes
of this second type, especially against their formal analysis, that
the attack is especially directed, in order to furnish, it is held, a
basis for the (erroneous) generalization that all analysis is false.
For, in the case of at least certain specific types of analysis here,
the terms to which analysis leads, if attention be directed exclu-
sively to them and the organizing relations be ignored, can be made
to seem to be the contradictory-opposite of the whole which is
analyzed. This contradictory character of the parts is sometimes
connected, though erroneously, with the application of the con-
stitutive theory of relations to the whole which is analyzed.1
However, that there is no empirical reason for applying this theory
to these wholes, and that analysis does not lead to contradictions,
if its results are stated accurately and justice is done it, will be
demonstrated in what follows.

The wholes of this second type are, as wholes, classes, or assem-
blages of similar individuals; they are designated by so-called
universals, by generic and abstract terms, terms with an exten-
sion and an intension. Examples of such wholes, so designated,
are carbon, American, even integer, etc. The class as many is
formed by the terms composing it. The terms themselves may
be certain other kinds of wholes, such as organic wholes, but they
are not classes, for that would give us the third type. The analy-
sis which leads to the discovery of the parts may be either ex-
perimental or formal, or both. For example, in the case of dis-
covering the atoms of any element, or of the atomic theory as
applied to any element, the analysis is, in part at least, experi-
mental ; perhaps in certain cases it is purely so, while in others,
such as that of discovering points, instants, irrational numbers,
derivatives, etc., it is formal. Formal also — and this is an im-
portant example — is the analysis of those changes, either of. posi-
tion or of state, in which the terms are identical with the deriva-
tives, -^ 's, of certain entities with respect to time.
ax

1 See the preceding section.

THE MODEL ATTACK 171

The actual analyses which are attacked, are, in the order in which
analysis shows they should be arranged according to the prin-
ciple of logical priority,1 those of number, space, time, change, mo-
tion, and change of state. Here, as with the first type of whole,
that attack which is based on the argument (Bradley 's), that any
and all diversity and so any and all relations of any and all terms
are self-contradictory, will not again be replied to, nor will certain
other positions, already dismissed. So also that attack which is
based on the arbitrary application of the constitutive theory of
relations will not be considered, for it is not made directly against
the analysis of this kind of whole — except by analogy. Only
that attack which is grounded on the supposed direct contradiction
between the terms revealed by analysis and the whole will concern
us here. The model for this attack is really, in every case, I
think, the specific attack on the analysis of motion, and this is
in its logic as old as Zeno. Briefly, this model attack goes as fol-
lows : Motion is given as a whole, to immediate experience. Its
chief character as a whole is that it is continuous change — of the
position of a body, in time. In accordance with the theory of
constitutive relations it might be held, and as a matter of fact is
held by Bergson, that what the analyst would attempt to distin-
guish as the body, the time, and the positions, respectively, are
fused and interpenetrated so as to form one whole. This is Berg-
son's doctrine of duration.2 But it does not enter this model at-
tack which we are now considering, and which Bergson also uses.
This is constructed by surreptitiously granting time to be the in-
dependent variable in the usual sense. Then the attack on the
analysis of motion follows. Analysis — so this model attack goes
— breaks space up into points, time into instants, and then makes
motion consist of rests. The moving body is in or at a point at
each instant, and this is interpreted as rest. But rest is the con-

1 See this essay, 205.

* Cf. Matter and Memory (Paul and Palmer, trans.), 1911, and Creative Evolution
9-10, 46, 201, 338-346, et passim. Bergson's doctrine of ' duration,' as an attack on
analysis, is examined later in this essay, 212.

172 DEFENSE OF ANALYSIS

tradictory of motion. Therefore does the analysis falsify. It
makes the originally given continuous entity, motion, a discon-
tinuous set of its own contradictories, rests. Each rest is thought
by itself, in fact when we think motion, intellectualize it, we make
it rests. Each rest is thought of as 'external' to the others —
as just that rest, identical with itself, unchangeable, and so the
very contradictory of a specific kind of change — motion. Each
rest is inert, and the various rests are external to and exclusive
of one another, and so are discontinuous with one another — at
least so the argument goes. On the basis, now, of this model at-
tack, the conclusion is reached, that not merely does this specific
analysis falsify, but that, quite analogously, all analysis ' spatializes/
works in terms of inertia, of geometry, of statics, of immobilities,
of discontinuous entities, etc. (surely a curious jumble which
a more careful analysis would clarify), and so falsifies that
which everything is, namely, a universal Becoming.1 But it is
clear that this model attack does not do justice to the actual analy-
sis against which it is directed, if, indeed, it states this correctly.
Clearly, it considers only the terms, and ignores the organizing
relations. But worse than this, it misstates the terms ; it sets up
a false analysis, only to knock it down — always an easy task.
For the analysis of motion, stated correctly, does not lead to rests!
If in this case, then, if in this model attack, justice is not rendered
analysis, and yet if the attacking party generalizes from this
model, does it not suggest that his entire attack is seriously
vitiated? My defense will confirm this suggestion.

The first kind of analysis which I shall thus examine is arith-
metical analysis, as it is found in authoritative sources, making no
claim to originality other than in the selection and arrangement
of results, and in some of the argumentation.2

1 Bergson in numerous places in the three works cited.

1 Among the works consulted here and in connection with the sections on space,
time, and motion are : Pierpont, J., The Theory of Functions of Real Variables,
1905 ; Dedekind, R., Essays on Number (Beman, W. W., trans.) 1901 ; Young, The
Theory of Sets of Points, 1906 ; Russell, Principles of Mathematics, 1903.

ARITHMETICAL ANALYSIS 173

1. Arithmetical Analysis. — Much analysis is intimately con-
nected with arithmetical analysis, for in this such general and im-
portant problems as those of continuity and of discontinuity, of
finitude and of infinity, etc., have received their most complete
and precise treatment. Here, however, only such features of
arithmetical analysis as must be taken account of in order to
defend it and to elucidate certain other analyses will be presented.

A. Numbers. — Modern arithmetical analysis finds good reasons
for grouping numbers into several classes ; there are ordinal and
cardinal numbers, positive and negative numbers, real numbers,
rational integers and fractions, and irrationals. Classifying these,
real number is the genus, and real numbers are either rational or
irrational, and, in turn, the rationals are either integers or frac-
tions.

B. Rational Positive Integers. — The rational positive integers
are the terms of the series, 1, 2, 3, 4, . . . n — 1, n, n + 1.

Some of the important characteristics of this series or whole
and of its terms are :

1. It is a series whose terms present an objective order of
magnitude, which is to be defined as a certain pair of indefinable
relations, greater and less, which are asymmetrical and transitive,1
each being the converse of the other. Terms capable of these re-
lations are magnitudes. Therefore the rational integers are magni-
tudes.

2. These numbers form a series or progression whose terms sub-
sist in a certain order by virtue of an asymmetrical transitive
relation such that, if x, y, and z be consecutive integers, and x < y,
and y < z, y = x-\-l, and z = y + 1, and xRy, yRz implies xRz.
The specific generating and organizing relation is, then, in this
case, asymmetrical and transitive.

3. It is a series or progression which not only lacks continuity,
but is not even dense or compact; that is, between any two of its
terms there is not another term or integer. It is thus discrete and

1 For definition, see below.

174 DEFENSE OF ANALYSIS

consists of terms which are only next to immediately preceding or
successive terms — a characteristic which is not found in the arith-
metical continuum.

4. The series is Archimedean; that is, there is no number a so
small but that some multiple of a, say na, is greater than any pre-
scribed positive number 6.1

5. The terms of this series are not existents; that is, they are
not of necessity correlated with specific instants of time or with
these and specific points of space. This means that they are not
psychological entities, although they can be correlated with these.
Counting does not generate, but, rather, presupposes them. Nor are
they physical entities. The assumption that they are presupposes
its contradictory.2 Indeed, both psychological and physical
entities imply these numbers, but not conversely. In this sense
the latter are logically prior to the former ; they subsist independ-
ently of all existential entities.3

6. Further, these numbers are neither spatial nor temporal;
that is, they are not terms either of finite (though small) spatial
extension in one or more dimensions, or of temporal extension, time.
The assumption that they are this presupposes the contradictory ;
for, hi defining and analyzing such extension or duration, not only
are elements of extension, points, which are not extended, and
elements of time, instants, which are not duration, implied, but
also the correlation of the numbers with these non-extended points
and instants is implied. Therefore the numbers are not extended.

7. Thus these numbers are logically prior also to both the points
of space and the instants of time. Instants and points imply the
integers, but not conversely. The integers, then, are individuals
or terms which are external to each other without being spatially
external. Externality, logically, does not have an* exclusively

1 This feature is important in connection with the later discussion of velocity
and acceleration.

1 If they were physical, they would be spatial and temporal ; but they are
neither. See 6 and 7.

1 See Perry, this volume.

TYPES OF RELATIONS 175

spatial connotation. Its ultimate logical meaning is exclusive in-
dividuality.

C. Relations. — Before going further in my statement of the
results of arithmetical analysis, clarity must be introduced in the
matter of relations. The term ' relation ' has already been used, but
not denned. Modern analysis tends to show, however, that ' rela-
tion' is an indefinable. Term, however, can be defined as any
entity which can stand in any relation, and terms can be classified
as physical and mental entities, complexes and simples, existents
and subsistents, classes, individuals, and relations.1 Leaving 're-
lation' undefined, however, analysis succeeds in distinguishing
and classifying relations. One of the classifications which is im-
portant for this essay is that which distinguishes relations as
symmetrical and asymmetrical, and as transitive and intransitive.

1. A relation is transitive if it is such that xRy, yRz together
imply xRz ; it is intransitive if this implication does not hold.

2. A relation is symmetrical if it is such that xRy implies yRx ; it
is asymmetrical if it is such that xRy excludes yRx. If xRy, then x
is referent, and y is relatum.

3. A relation is ordinally correlating if it is such that it couples
every term of a series s with a specific term of another series s',
and vice versa, in the same order in each. It is merely correlating
if it couples any term of the one series with any term of the other
in any order. In either case it is one-one.

4. A relation is one-one when, if x differs from x', and y from
y', there subsist xRy and x'Ry', but not xRy' and x'Ry.2 Motion
is a specific instance of such a one-one correlation between points
and instants.

5. A relation is many-one if, x and x' being different, xRy and
x'Ry both subsist. Rest is a specific instance of a many-one cor-
relation between many instants and one point.

6. In these cases, moreover, if x, the referent, belongs to some

1 Not as relating, but as related in another context.
* Cf. Russell, Principles of Mathematics, 113 and 305.

176 DEFENSE OF ANALYSIS

class contained in the domain of the relation, then the relation de-
fines y as a function of x, and the series of x's forms the independent
variable, the 'argument,' and the series of y's the dependent vari-
able; that is, an independent variable is constituted by a series of
terms each of which can be referent to a certain relatum. For ex-
ample, time is the independent variable in both rest and motion.
In actually existing rests we have one point correlated with many
instants through the occupation, by a material particle, of that
point for a finite time. In motion each point is correlated with
one and only one instant of time through the occupation, by a
material particle, of each point for an instant. But this occupa-
tion is not to be interpreted as rest.

7. There are two theories of relations, known respectively as
the theories of external and of internal relations.1

This presentation of the results of analysis in the field of the
integral numbers suffices to make clear what the reply must be to
any attack on this specific kind of analysis. Clearly in the case of
the rational integers there is given no whole which is prior to the
parts, and which, as a whole, can be contrasted with the parts to
prove the falsifying character of the analysis. The parts cannot
be shown to be the contradictory of the whole, because there is
no whole which is empirically given prior to them. The only
whole is the series or class of individuals in a certain relation.
Any other kind of whole here could be constructed only on the
basis of arbitrarily applying the theory of internal relations to the
integer series. But this would be a very artificial procedure, and
is, in fact, carried out by no one. We may conclude, then, that
the analysis of the integer series, as it has been stated hi its essen-
tial details, is quite sufficient to present the character of that whole
and of its analysis, and to refute that type of attack which claims
that such analysis is falsification, that numbers are spatial, etc.
In fact, this analysis offers no opportunity for the argument that
analysis leads to parts which are the contradictory of the whole.

1 This essay, Section II. Asymmetrical relations are unintelligible on any other
theory than that of external relations. See Russell, ibid., XXVI.

INTEGERS AND RATIONAL FRACTIONS 177

D. Rational Fractions. — So far as the purpose of this essay
demands that the analysis which deals with the rational fractions
shall be considered, these numbers are treated in the subsequent
section on the whole group of rationals. Suffice it to say, that,
leaving division and multiple undefined, a rational fraction is that
number which subsists by virtue of any integer a being a multiple
of any integer 6, and which is represented by the customary sym-
bols a/6 or a : 6.

E. The Whole Group of Rationals, both Integers and Fractions. —
Analysis reveals the following important (for this essay) char-
acteristics of this assemblage of rational integers and rational
fractions.

1. The terms of this assemblage form as many types of series
as they are objectively capable of; that is, there are as many
orders among them as there are defining serial relations of which
the terms of this assemblage are the field. The order of magni-
tude is one such order, the so-called 'natural order.'

2. Any one of these orders is a series or progression whose terms
subsist in some specific order by virtue of an asymmetrical tran-
sitive relation.

3. A proper name can be given to each of the terms of any
series of this assemblage.

4. Any of the series of this assemblage is dense or compact; that
is, between any two terms there is an infinite number of terms of
the same series.

Definition. By an infinite number is meant such a class, u,
that u — 1 (class u') is in one-one correspondence with u ; u is
then similar to a proper part, u', of itself.

5. By virtue of 4 there is, in any series of the assemblage,
no term immediately preceding or succeeding any specific term ;
that is, no term is next to any other. Yet. no series of this as-
semblage is continuous in the strict sense of this term. The
assemblage is only compact. For example, it does not subsist in
one-one correspondence with all the points of a right line.

178 DEFENSE OF ANALYSIS

6. The assemblage is Archimedean.

7. The general concluding statements (5, 6, and 7) which were
made above concerning the nature of the cardinal integers, their
relation to psychical and physical entities, to counting, etc., hold
good also for this assemblage and need not be repeated here.

F. Irrational and Real Numbers. — That there are irrationals
is discovered in the realization that there is some value for x,
whereby, for example, x2 = 2. This value is found to be neither
a rational integer nor a rational fraction. It belongs to that class
of numbers which are called irrationals. What these numbers are
cannot in the nature of the case be stated by the proper names
which are applied to the assemblage of rationals, but it is possible
to determine an infinite sequence of rational numbers, a\, 02, as . . .
such that each number, on, satisfies more nearly than the preceding
ones the conditions of the problem.1

Irrationals may accordingly be defined as the limits of such
series of rationals as have neither a rational nor an infinite limit,
a limit being defined as follows: "L is the limit of the sequence
A = (aB) when for each positive rational number e, chosen small
at pleasure, there subsists an index ra such that I — an <«." 2

The importance of the irrational numbers rests in part upon
the fact that only through them is the assemblage of numbers,
that is, integers, rational fractions and irrationals, continuous in the
strictest sense of this term. This is one of the principal reasons why
the presentation of arithmetical analysis has been introduced into
this essay. For one of the chief attacks on analysis is that which
holds that analysis falsifies that which is continuous. It is im-
portant, then, to present clearly and with precision that which
analysis shows the continuum to be. We shall then be in a posi-
tion to do analysis justice, and to find out if it really does lead to
contradictory parts and so falsify. The nature of the number
continuum may be stated as follows:

Let t] be the assemblage of rational numbers, both integers and

» Cf. Pierpont, op. cit., 35. » Pierpont, ibid., 25.

IRRATIONAL AND REAL NUMBERS 179

fractions, in order of magnitude. This series is denumerable
and compact, as has been seen. Let 6, now, be the assemblage
of rationals and irrationals. Then 17 belongs to 0. ®, then, is of
such a character that between any two of its terms there are terms
of rj} and conversely. Then the assemblage of numbers constitut-
ing 0, in order of magnitude, is the number continuum.1 The fol-
lowing are the attributes of this assemblage :

1. Between any two assigned limits, say, 0 and 1, there are all
values or numbers.

2. The assemblage includes all its limits and all its members
are limits. (See previous definition.)

3. These members or terms are individuals, infinite in number,
external to each other, each different from every other, and are
related by an asymmetrical, transitive relation.

4. These individuals are not analyzable into new terms like to
or different from themselves, although the proper names by which
some of them are designated may be. They are simples, not
complexes. The assumption that they are so analyzable repeats
the problem, and is ultimately found to presuppose its contra-
dictory.

5. The assemblage is perfect2 in that it contains its first de-
rivatives, -^ 's, or limits. Every term is a limit or derivative, and

dx

this is important since thereby a one-one correspondence subsists
between this series and the points of a graph representing motion
of uniform velocity, or the change of velocity, etc. For then each
term of the assemblage, between assigned finite limits, is the value
or limit of a certain ratio or relation between, for example, a cer-
tain finite change of position on the one hand and a finite temporal
period on the other, and these limits form a continuous series.

6. The assemblage is Archimedean.

7. If « (any real number) > 0, there is an infinity of rational
numbers, also of irrationals, < «, and also > a .

1 Cantor, Math. Annalen, 1895, 46, 481.

* Cf. Pierpont, op. cit., 162 and 168, and Russell, op. tit., 291 and 342.

180 DEFENSE OF ANALYSIS

8. Between a, ft, any two real numbers, « < ft, there is an
infinity of rational numbers, also of irrationals. Then the assem-
blage is continuous in the strictest sense of this term, and there
are no next or consecutive numbers. For assume that «, y are
such ; then between them there is an infinity of real numbers, and
a, y are not consecutive.

The further general statements which have been made with
reference to the integers and the rationals may also be made with
reference to the assemblage of real numbers. Thus,

9. They are not existents, but are subsistents; that is, they are
entities which are quite independent of all existents, both physical
and mental, although the converse proposition does not hold.
Existents, and space, and time imply them, but not conversely.
Nor are they, of course, spatial or temporal entities. They are
logically prior to space and time, although they are correlated
with points and instants.

As concerns the possible attack, now, it is with the real num-
bers as it is with the assemblage of rationals. There is no whole
which is given empirically either prior or subsequent to the parts
which analysis reveals. The only whole is that which is identical
with the class of entities denoted by the concept, real number, and
which consists of the individuals composing it. There is no whole
with which the parts can be contrasted in order to reveal a con-
tradiction and so the falsifying character of the analysis. Then
there is no opportunity for an attack on this analysis except from
the standpoint of a whole which may be constructed artificially by
the use of the theory of internal relations.1 But such a procedure
presupposes the very analysis it would invalidate.

The entire character of the assemblage is revealed by that analy-
sis, some of whose mam features have been presented above, and
which discloses not only terms, but also relations. This analysis
presents the most exact definition of continuity extant, one which,
while it demands the complete externality of terms and the theory
of external relations, precludes the interpretation of this external-

1 See this essay, Section II.

THE ANALYSIS OF SPACE 181

ity as spatial or temporal. The real numbers are not spatial or
extended, nor have they space between them, and, although,
psychologically, knowledge of them may depend upon the knowl-
edge of spatial things, logically they are not dependent on space.

2. The Analysis of Space. — We are now in a position to examine
understandingly and with precision some of the important facts
concerning space which have been revealed by modern analysis.
These facts enable us again to confute that specific attack which
we have found to be the most formidable. Reasons have already
been given for excluding from consideration the other types of
attack, such as the pragmatic, and the Bradleian. The attack
which I am meeting proceeds for the most part along the lines of
the model presented on a previous page.1 This attack, on model
lines, runs as follows : Space is given empirically as a one-, or two-,
or three-dimensional continuous whole ; or it is given as a whole
which analysis resolves into dimensions. This analysis into di-
mensions is, however, not attacked. Now a dimension, or a spatial
extension in one dimension, is analyzed into points. But a point is
unextended, it has no dimensions. It is the contradictory of the
whole, the extended 'thing.' Either, then, this analysis is false, or,
if any analysis of the line is to be accepted as true, it must be that
analysis which does not lead to points, but to elements having ex-
tension in at least one dimension.2 For to derive extension from
the unextended is impossible. Either no analysis, or only that
analysis which is division and leads to extended elements — though
these be very small ! This is the dilemma, and this is the way the
argument runs.

Is the analyst in a position to reply to this attack effectively,
and can it be shown that the attacking party does not do justice
to the actual results of analysis, or that he misstates these results,
or himself analyzes falsely? The answer is 'yes' in each case.
Let us consider some of the actual results of this analysis, noting,

»171.

* Cf. Kant's second antinomy, and Zeno's historical first and fourth arguments.

182 DEFENSE OF ANALYSIS

however, that the question of the objective or (Kantian-) sub-
jective character of space is not here involved.

The modern general analysis of space finds reasons, first, for
distinguishing three more specific analyses or geometries, pro-
jective, descriptive, and metrical.1 Yet all three of these geome-
tries agree in analyzing space into points; constructively they
are said to assume the class-concept, point. The actual points
are the individuals which this concept denotes; they are its ex-
tension.

Point is, perhaps, indefinable. Yet it is found to have all the
characteristics which the real numbers have, plus something more.
It has a peculiar quale which can at best be defined only in terms of
that of which it is the element, namely, space. But that may be
to define the term in a circle and to admit it to be indefinable
logically.

Constructively, now, each of these three geometries assumes
two points. These two points determine a class, the straight line,
analogously to any two real numbers determining the series of real
numbers. Then there are at least three points, A, B, and C, and
either A is between B and C, or B is between A and C, or C is be-
tween A and B. The further assumption is made that there is
one point not of or on any given straight line; that is, that there is
a fourth point, D. These four points determine a class, the plane.
Assume, next, a point G, not of this plane. Then G and any three
of the class of points which determine a plane determine a space
of three dimensions. Space, then, is a class which is defined by
its relation to the class-concept point. It is the domain of the
concept, point, and of a certain class of relations holding between
points.

What, next, are some of these relations? Modern analysis
shows one of the most important of them to be that which is ex-
pressed by the word between. The definition, giving the very
meaning and the criterion of betweenness, is as follows : "A term
y is between two terms x and z with reference to a transitive asym-
> Cf. Russell, op. c&, XLV-XLVIII.

THE ANALYSIS OF SPACE 183

metrical relation R when xRy and yRz." l Between, then, is a
relation which determines an order of sequence of the points of a
straight line, of a plane, of space. This relation receives a specific
and precise statement in the following propositions which I quote
from Hilbert's "Foundations of Geometry."2

1. If A, B, C are points of a straight line and B lies between A
and C, then B lies also between C and A.

2. If A and C are two points of a straight line, then there exists
at least one point B lying between A and C and at least one point
D so situated that C lies between A and D.

3. Of any three points situated on a straight line there is always
one and only one which lies between the other two.

4. Any four points A, B, C, D of a straight line can always be
so arranged that B shall lie between A and C and also between A
and D, and, furthermore, so that C shall lie between A and D and
also between B and D.

It results from these four propositions that between any two
points of a straight line there subsists an infinite number of points.
Then in this respect the line resembles the series of rational frac-
tions, and also of real numbers. But it is further shown thereby
that the defining relations which relate the points are similar to
those which relate the integers, the rational fractions, and the
irrationals; that is, they are asymmetrical, transitive relations.
Thus, any one of the above propositions implies that, if x and y
are any two points of a straight line, then xRy excludes yRx. For
example, by 2, if x and y are two points of a line, then there subsists
a third point z lying between x and y, and at least one point q
such that z is between x and q. This may, perhaps, be more
readily seen if we state the axiom of Archimedes as given by Hil-
bert, "Let x\ be any point upon a straight line between the arbi-
trarily chosen points x and y. Take the points x2, z3, z4, . . ., so
that x\ lies between x and xz, x% between x\ and xs, x3 between z2
and z4, etc. Moreover, let the segments, xx\, x\x%, x^x^ x$x\ . . .,

1 Russell, ibid., 214.

* HUbert, D., (Townsend, E. J., trans.), 6.

184 DEFENSE OF ANALYSIS

be equal to one another. Then, among this series of points there
always exists a certain point xn such that y lies between x and xn." l
This gives a definition of the continuity of the line.

But further, if the numbers by which we designate the points of
the line are the real numbers, and these numbers are taken in their
order of magnitude, then the points which they designate must also
have this order. But asymmetrical, transitive relations hold be-
tween these numbers in this order. Therefore they hold between
the points.

The necessity of introducing the real numbers in order to desig-
nate all the points of a line is shown by examining, for example, the
relation of the hypotenuse of a right isosceles triangle to the two
sides. Let these, metrically, be of the value 1 ; then the hypot-
enuse = V2. But this is an irrational. Therefore some of the
points of a line are in one-one correspondence only with the irra-
tionals. It is alone all the real numbers, that is, the rationals and
the irrationals, that are in complete one-one correspondence with
all the points of a line.

Space, then, is shown by analysis to have the same kind of con-
tinuity (and also infinity) that the real number series has. But
its continuity is a continuity of points, not of numbers; that is,
there is a quale in the spatial element, the point, which the num-
bers lack. But there are also common properties. The points,
like the numbers, are individuals, simples, not further divisible
into similar elements, and are logically external to one another.
The defining relations are also the same. They are asymmetrical
and transitive. This logical externality of the members of this
particular class of individuals and the asymmetrical, transitive
relations holding between them are space of one, or two, or three
dimensions, are spatial externality. Quite similarly, time is the
class of individuals, called instants, related asymmetrically and
transitively, external to one another, and not further divisible into
similar elements. These individuals hi turn have a quale different
from the quale of the point.

lOp. tit., 25.

CONTINUITY OF SPACE 185

Accordingly the ' axiom of the continuity ' of space may be stated,
in agreement with the above and preceding discussions, as fol-
lows : All points on a line are limits of rational points, and all
infinite series of rational points have limits.1 By this definition
all the points of a line form a perfect series ; for a series is perfect

if all its limiting points (derivatives, -^ ) belong to it, and if all its
\ dxj

derivatives are limits. But this is the case with the points of a
line. A series is continuous, then, if it is perfect. But the series
of points is perfect. Therefore it is continuous.

Analysis shows, further, that descriptive space is in its continuity
like projective, and that the above axiom satisfies also the con-
tinuity of metrical space.2 In general, the above results show the
way in which modern analysis treats of space, of some of its char-
acteristics, of its elements. This analysis states with clearness
and precision what space is, what its continuity is, what terms and
relations are involved. Space or continuous spatial extension is
the domain of an asymmetrical transitive relation between points.
The continuity and extension are defined and determined by ele-
ments and relations neither of which are, strictly speaking, con-
tinuous or discontinuous, and, possibly, either extended or unex-
tended.3 Analysis shows that the fundamental relations essential
to geometry do not hold between spatially divisible terms, such as
lines, planes, and volumes, but only between spatially indivisible
points. It shows that however many parts a dimension be divided
into, the parts are still dimensions or stretches with which many
numbers are in one-one correspondence. But there are single
numbers, simples. Then, corresponding to these, there must be
single points, simples, not stretches, and not further analyzable
into elements of the same kind. Yet the continuity of space is
defined and determined not alone by these elements, but also by

1 Cantor, G., Ada Mathematica, 2, 341-344 and 405 et passim.
» Russell, op. tit., XLV-XLVIII.

1 They might be eaid to be non-extended, even as inanimate objects are non-
moral.

186 DEFENSE OF ANALYSIS

the defining relations between them. Let these be the premises.
Then from them, from the assumption of points in a certain rela-
tion, a space empirically indistinguishable from the space in which
we live can be constructed. Observed facts are explained by our
premises. Thus the latter have empirical confirmation.

These are the results which are attacked by the critic of analy-
sis — by him who would establish its falsifying character. When
justice is done them, when the actual analysis is considered, can
the attack succeed?

The attack is made along two lines, but in both cases it follows
the model previously presented. In one attack the continuity-
feature of space is selected as the fulcrum of the argument. Space,
the attacking party holds, is given empirically by intuition, or some
such mode of direct approach, as a unitary continuous whole.
But analysis leads to terms or to parts of space which are discrete
from one another. Therefore analysis leads from continuity to
discontinuity. It leads to the contradictory of the thing analyzed.
Therefore it falsifies.

The other attack proceeds somewhat differently. The analy-
sis of space, it holds, leads to terms which are not spatial ; it leads
from the extended, the dimensional, to the unextended, the 'un-
dimensional.' Then, again, it leads from the positive term to the
contradictory. Therefore it falsifies.

Manifestly, in both cases, the attacking party considers only the
terms and quite ignores the relations which analysis also discovers.
Clearly he does not do analysis justice. Indeed, he may not even
be defining the terms correctly in asserting them to be unextended
and undimensional. They may be simply non-dimensional.

Can these attacks be repulsed ? The answer can again be only
affirmative, and, in fact, has already been given. The first requisite
condition for determining whether analysis is intrinsically false or
not, is that justice shall be done the analysis. Then, not only
terms, but also organizing relations must be considered, and false
analyses must not be set up, only to be knocked down. Observing
this condition, we may reply to the first attack as follows :

ATTACK AND REPLY 187

In the case of space, as distinct from the numbers, there does
seem to be a whole which is given psychologically, to any individual
consciousness, prior to the parts which analysis reveals, and this
whole may be said roughly to be continuous and infinite, etc. But
to neither of these terms is therewith an exact meaning given.
Analysis endeavors among other things to give just this requisite
exactness, to make precise that which is otherwise a vague mass
of confusion. The first attack selects division as that phase of
analysis which it would impugn. Division is held, or made to
mean discreteness. But this is just what the analyst finds it does
not mean. Division in the usual, in fact in any strict sense of the
term, does not lead to entities which differ from the entity divided,
whether this be space or time, or space traversed in time, or what-
not.1 The parts resulting are in turn spaces, or times, or motions,
and never points, or instants, or velocities. Though very small,
they are always finite quantities ; they are stretches, and never
limits. But the very condition for there being such a division, for
there being such a 'thing' as the so-called discreteness which the
attacking party talks about, is that which is really identical with
continuity itself. Clearly that which might determine the dis-
creteness of two segments of a line need not itself be a segment.
It could be a point. But that every point of a line generates a
'partition' of the line, and that every 'partition' is generated by
a point is precisely Dedekind's definition of continuity.2 The
same 'partition' definition holds also, of course, for the numbers.
Further, the point is not a third entity, which, film-like, belongs
to neither segment, but it belongs to either one or the other of the
two segments which it determines, at the same time that the line
is continuous. "Thus, " by Dedekind's definition, " if all the points
of a line be divided into two classes of which one precedes the other,

1 Cf. this section, on Motion.

1 Dedekind, Essays on Number, 20. IV. "If the system R of all real numbers
breaks up into two classes U\, t/a such that every number ai of the class U\ is lesa
than every number 02 of the class Us, then there exists one and only one number a
by which this separation is produced." Cf. Pierpont, op. cit., 78-79, and Russell,
op. cit., 438.

188 DEFENSE OF ANALYSIS

then either the first class has a last term, or the last a first, but both
do not happen."1 We must conclude, therefore, that this first
attack on the analysis of space, which makes analysis lead from
continuity to discreteness, fails of its purpose. The discreteness
which it would introduce proves to be only another definition for
the continuity which the discreteness is supposed to displace.

The second attack on the analysis of space is based on the con-
tradiction which is supposed to hold between the elements to
which analysis leads and the original entity analyzed. The entity
analyzed is extension in one or more dimensions. The element
discovered by analysis is unextended. Thence the contradiction,
and the falsifying character of the analysis ! Or, it is maintained,
the entity analyzed is continuous. But the elements discovered,
the points, are individual, and so, separate, discrete, discontinuous
from one another. Manifestly neither of these arguments does
justice to the actual analysis. Let not only the terms, but also
the defining relations be considered, and the attack in each case
is nullified. Both cases may be taken together. It is con-
tinuous extension which is to be analyzed. And analysis shows
that this is identical with a series of points, each individual, and
so logically external to the others, related asymmetrically and
transitively. Extension is just this, a series of points, with each
point distinct from every other, and any two, A and B, implying
a third, C, beyond B, so that B is between A and C. What else
could spatial extension be than just this field of an asymmetrical
transitive relation holding between points? And spatial con-
tinuity is just this too, a series of points related in such a way that
between any two there is an infinite number of points in one-one
correspondence with all the real numbers. To grant the second
aspect of the attack, that the points are individual and discrete,
either nullifies the attack by making this very discreteness identi-
cal with continuity as above defined, or repeats the problem, if the
discreteness of the points is interpreted to mean extension between
them. For then this extension must be analyzed, and sooner or

1 Russell, loc. cit. Cf. Dedekind, op. cit., 11.

CONCLUSION ON SPACE 189

later the recognition will come that this involves points related in
a certain way, and that there is no escape from this — that points
so related are one-dimensional continuous extension.

We may conclude in general, therefore, that the modern analysis
of space stands quite unimpugned by the type of attack considered
in this essay, and also, since arguments have been presented else-
where which are held to refute other readjustments and reinter-
pretations, that its results must be interpreted realistically.
Points and the specific relations between them subsist quite in-
dependently of the knowing or discovery of them. Only one type
of attack has any plausibility at all, namely, that which makes
analysis lead to the contradictory of the thing analyzed. But this
attack does not do justice to the actual analysis made. It selects
the terms, the points, and neglects the relations. Consider these,
as well as the terms, and the contradiction disappears, and the
attack is invalidated. 'Analysis may lead to terms which are dif-
ferent from the thing analyzed, and this difference can always be
thrown into the form of a contradiction.1 If A differs from B, B
can be called non-A. Were this fact damning, it would damn the
very method by which it is discovered. But that it is not damn-
ing, and that distinctions can be made which, though convertible
into contradictory form, are nevertheless valid, is presupposed,
if any analysis whatsoever is made. But the attacking party al-
ways uses analysis. If not damning in one place, then it is not in
another, until evidence to the contrary be adduced. And so with
the analysis of space. Let its elements be points, different from
spatial extension. But consider also the relations. Then space is,
both in its continuity and its extension, the assemblage of these
elements related asymmetrically and transitively, and the analysis
is adequate. Space is just this assemblage. There is nothing left
over, as characterizing the whole, which is not these elements in
these relations.2

1 Cf. Holt, this volume.

2 This outcome and the considerations leading up to it solve Kant's second
antinomy.

190 DEFENSE OF ANALYSIS

3. The Analysis of Time. — At this point I shall consider only
those aspects of the analysis of time which might be subjected to
the typical attack by means of the argument from contradiction.
There is another attack which I shall consider later, that, namely,
which is identical with the argument that the analysis of time is
contradictory to its very nature as lived, — as 'duration.'1 This
last argument is based on the constitutive theory of relations.
But at this point we are concerned only with the attack which is
directed against an analysis proceeding along quite the same lines
as does the analysis of space just presented. Since the two analyses
have much in common, except that here we have to do with a one-
dimensional manifold of instants, the statement of the results of
this modern analysis of time may be made very brief. It is as-
sumed here, since it is justified later, that the analytical separation
of time from things in time, like the analytical separation of space
from things in space, is valid.

Modern analysis shows that time as such is of the following
character :

1. Time is a one-dimensional infinite and continuous manifold
of individuals, called instants, related by an asymmetrical transi-
tive relation.

2. Instant is, perhaps, indefinable. Instants have all the
characteristics which the real numbers have, and more besides.
An instant has or is a peculiar quale which is different from the
point quale, and is an 'addition' to the real number quale. This
peculiar quale can at best be defined only in terms of that of which
it is the element, namely, time.

3. The element of space is the point, and space is either a one-,
or a two-, or a three-, or an n-dimensional series of points. The
reason for so saying is, perhaps, that empirical space seems to be
three-dimensional. So, empirically, time seems to be given as
only one-dimensional, but logically time may be, like space, n-
dimensional.

4. A one-dimensional manifold or series may be defined as

1 See this bection, on Dynamics and Duration.

THE ANALYSIS OF TIME 191

follows : Let there be a transitive asymmetrical relation R and
a collection of terms any two of which are such that either xRy or
yRx. Then, since by definition R is asymmetrical, xRy is differ-
ent from and excludes yRx ; and, since R is transitive, xRy, yRz
imply xRz. Then R, the converse relation, is also asymmetrical
and transitive. It results that any three terms of the series are
such that one is between the other two, and the whole collection is
a single series.

5. A series of two dimensions subsists, if every term u\ of a
series of which x, y, and z are some terms, is itself an asymmetri-
cal transitive relation which generates a series. The class Uz of
terms forming the field of all the relations in the series generated
by R is a two-dimensional series. A series of three-dimensions
subsists if each term of u2 is itself an asymmetrical transitive
relation generating a series. In quite a similar way a series of any
number of dimensions can be defined.1

6. Time, therefore, for empirical, though not for logical reasons,
may be said to be a one-dimensional series of instants.

7. The time series, then, has the same properties as has the one-
dimensional space series with the difference that in the former the
terms are points, in the latter instants. Accordingly the prop-
ositions2 which were made above concerning the space series
hold good of the time series, if ' instant ' be substituted for ' point.'

8. It results from the four propositions thus made, that between
any two instants there subsists an infinite number of instants. In
this respect, then, the time series is like the series of real numbers
and the space series. A one-one correspondence subsists between
the real numbers, the instants of time, and the points of space.

9. In just this respect, therefore, and as thus defined, time is
continuous. It is continuous in the same sense as is the number
continuum formed by the real numbers in their natural order; that
is, it is continuous in the most precise meaning of the term con-
tinuity. But such continuity subsists only by virtue of both terms
and relations. The terms, instants, are individuals, simples, not

1 Cf. Russell, op. cti., 374-378. 'This section, 183.

192 DEFENSE OF ANALYSIS

further analyzable into terms of the same kind, and external (not
spatially) to one another. This logical externality is identical, on
the one hand, with the distinctness of the terms, and on the other
with the property that, if A and C are two instants of time,
there subsists at least one instant B between A and C, and also at
least one instant D, so that C is between A and D ; briefly, there
is another instant D ' beyond ' C.

Time, or continuous temporal extension, is, therefore, the domain
of an asymmetrical transitive relation, the terms of which are in-
stants. The instants are themselves neither continuous nor dis-
continuous.

The attack on the analysis of time, so far as it follows the
model attack, fails for the same reason that it fails when directed
against the analysis of space. And the attack with which we are here
concerned does follow this model attack. It grants that time may
be validly distinguished from things in time,1 but holds that even
so it is given empirically, by intuition,*etc., as a unitary continuous
whole. Then it further holds that the analysis of this whole leads
either to instants which, as not durations or time extension, are
the contradictory of the whole, or to instants discrete and dis-
continuous from one another, which discreteness is contradictory
to the continuous character of the whole.

Can this argument be refuted ? Yes, if justice be rendered the
actual analysis, and the relations as well as the terms be considered.
The terms by themselves may seem to be the contradictory of the
originally given whole. But actually they are terms in a certain
relation. As terms in this relation, they present no contradiction
with the properties of the whole. In fact, only through them as
terms in relation is the whole what it is, — continuous, infinite,
extended — unless these attributes be left wholly vague and
undefined. The principle of indirect proof may be used to show
this. Thus, to assume that the instants are temporally extended,
or that there is temporal extension between them, or both,

i Cf., e.g., Bergson, Creative Evolution, 46 and 21, 22, 37, 39, 321-337, 341-345
and various other places.

ATTACK AND REPLY 193

but repeats the problem and so presents only a new necessity for
analysis, and reveals the fact that ultimately temporally non-ex-
tended terms, instants, in an asymmetrical transitive relation,
are presupposed. Time is just these terms in this relation, and
there is no characteristic of the empirically given whole over and
above what these terms in relation are.

Likewise the attack on the second basis fails. The discreteness
which analysis or division is held to introduce is exactly the kind
of discreteness which is identical with continuity, namely, that
whereby the formal 'partition' of a series by any term is made.1
It is not the discreteness of the integers, in which certain terms
are next to certain others, with no terms between, but it is the
discreteness of the real numbers, by which an infinite number of
terms is between any two and any term makes a 'partition.'
That is, the very condition for that formal division into segments
which the attacking party holds to mean the introduction of
discreteness into that which is not discrete, is that which makes or
is continuity in the precise and exact sense of the term.

We may conclude that the attack on the analysis of time
qua time fails of its purpose completely. Other attacks, such
as the pragmatic and the Bradleian are again excluded from
consideration here for reasons stated elsewhere. The realistic
interpretation of the results of the analysis of time will stand,
therefore, if no other attack can be brought against it and succeed.
Phenomenalism and idealism make such an attack or interpretation,
of course, but these are refuted by the arguments for the general
realistic position. There is another specific attack, involved in
Bergson's doctrine of duration, but this, I shall presently show,
presupposes the very realistic interpretation which it is intended
to displace.2

4- Motion and its Analysis.3 — The attack on the modern an-

1 This section, 187. * This section, on Dynamics and Duration.

3 In this section I am largely indebted to Mr. Bertrand Russell's analysis of
motion, contained in Chap. LIV of his Principles, though I depart from his views in
certain essentials.
O

194 DEFENSE OF ANALYSIS

alysis of motion proceeds as follows : First, it is held that any
finite motion is given as a continuous whole, definable only in terms
of itself. The determination of any specific instance of motion as
finite may be arbitrary, or it may not be. If it is, then it is
held to be continuous over or through the arbitrarily set limits,
and to be a unitary whole, unanalyzable and indivisible1 in its
character as motion, through all such limits. However, various
aspects of the motion, it is admitted even by the attacking
party, can be distinguished and determined by analysis, by
science. Its direction or path, its velocity or rate, its constancy
or its change of rate, are such aspects. The attack is not
directed against the analysis which discovers these distinctions.
Now the argument of the attack is as old as Zeno, and yet it is
advanced against analysis by Bergson,2 and with the seem-
ingly willful ignoring of the advance which modern analysis has
made since Zeno. Actual motion as experienced is a continuous
whole between either arbitrary or natural limits A and B. It
requires a certain time, and in that time a material particle or body
travels from A to B. It involves space, then, as well as time.
But the space and the time are composed of points and of instants
respectively. Then the particle or its center is in or at a point at
each instant. But this means that it is at rest at each instant.
Rest, however, is the contradictory-opposite of motion. Then
that analysis which leads from motion, the whole, to rests, the
parts, falsifies. It is not to be accepted as giving realities, but,
at best, as being only an instrument which serves our human
purposes and action.3

This is, as we have seen, the typical attack. It can be ex-
tended, in a way analogous to the attack on the analysis of
space and of time, to include the contradiction between the con-
tinuity of the original motion and the discontinuity (that which

1 E.g., Bergson, Creative Evolution, 304-313.
* Ibid., 163.

1 Cf., e.g., Bergson, op. ctt., Index, under Intellect and Action, for reference to
a manifold of statements to this effect.

THE ANALYSIS OF MOTION 195

is held to be such) between the rests, and perhaps there are
other extensions or variations. But all can be replied to and
refuted, if the actual results of the modern analysis of motion be
considered in their fullness, and be not misstated.

What are these results? The preceding discussions have pre-
pared the way for their presentation and explanation, so that
the precise problem now at issue is whether an exact, a non-fal-
sifying analysis of motion as motion, and a definition of motion in
terms of that analysis, can be made.

Now such an analysis and definition do not logically involve
direction, or rate, or even change or constancy of rate, or cause,
although no existential motion can take place without these
characteristics, or be empirically discovered without some of them.
However, once discovered, it can be analyzed and defined without
implying them. That this is the case can be made clear as follows.

To determine empirically that there is motion demands the
observation of two points, A and B, the possible specification of
these two points by three coordinates as a frame of reference,
and the observation, directly or indirectly, of the traveling of a
particle or body from A to B. But with this specification the
general direction of the motion is also specified. By general obser-
vation, its path, its course or curve can also be discovered, while,
by means of applied mechanics and further observation, the
equation of its curve, its constant or its changing velocity, etc., can
be determined. But with all these determinations made, and with
some of them necessarily involved hi the empirical ascertainment of
the fact of motion, the motion can nevertheless subsequently be
defined and analyzed without involving them, and even without
involving anything such as a material particle moving.

So defined, motion is the series of individual one-one relations
correlating the terms of two series, the time and the space, in such
a way that, if x and x' be any two terms of the time series, y and y'
any two terms of the space series, and x has the correlating relation
to y, that is, xRy, then it does not have this relation to y', nor does
x' have it to y. In a manner, then, somewhat analogous to that in

196 DEFENSE OF ANALYSIS

which a two-dimensional spatial series is the domain of an asym-
metrical transitive relation R between asymmetrical transitive re-
lations between points,1 motion is a series of complex terms, each
of which consists of a one-one correlating relation between
a specific term of the space and a specific term of the time
series; and just as the terms of each of these two series are
related asymmetrically and transitively, so are these complex
terms and the correlating relations similarly related. In sym-
bols, using r for the correlating relation, R for the asymmet-

\xRx'Rx" .

rical transitive relation, we have

rRr'Rr"

in which

yRy'Ry" . .

xRx'Rx" is the time series, yRy'Ry" the space series, and the
whole is motion.2 But further, just as the time and the space series
are each continuous, so is the motion series, and all the positive
statements which in the preceding sections were made with regard
to the continuity of the first two series hold good of the motion
series.

Motion is, therefore, a series, the domain of a relation which relates
relations which correlate terms in a one-one manner. Briefly, it is a
series of complex terms, xRy or pRi (p = point, i = instant, and r =
the one-one correlating relation). But it is not the complex term,
xry, or x'r'y', etc., that is, the correlating relation with its correlated
terms x and y, or x' and y', etc., of the two series, the time and space.
Nor is this complex term rest, for rest is logically analogous to mo-
tion. Rest is or involves a many-one correlating relation R between
many instants of time and one point of space hi such a way that
xRy, x'Ry, x, xr, x" . . . being different and the T/'S identical.
Quite analogously too, a logical definition of impenetrability and so
of matter can be given.8

Impenetrability is or involves a many-one correlating relation
R between many points of space y, y', y" . . . and one instant of

1 See 191.

1 A term can stand in many relations. Cf. Perry, this volume, Section III.

1 See Russell, op. tit., 467 and 480.

MOTION AND REST 197

time x, so that xRy, xRy', xRy", are different and mutually
exclusive. Although, now, the correlating relations in each one of
these three cases are similar, nevertheless they are individual and
in this sense different, and that they must be this is implied by the
individuality of x, x', x" . . . and of y, y' y" . . . respectively.1

Rest, then, is also a series of complex terms, namely, of correlat-
ing relations with their terms, themselves related asymmetrically
and transitively, and likewise with impenetrability. Motion, rest,
and impenetrability are each continuous. The possibly sur-
prising character of this conclusion is rendered less so by the
realization that neither motion nor rest is either space or time,
but that each must be in some way a matter of relation between
these entities.

However, these definitions and analyses so far are purely kine-
matical; that is, they concern simply the geometry and arithmetic
of motion. They do not involve actual matter, or causation, or
velocity, or acceleration, or direction in any other sense than that
an asymmetrical relation has or is direction. When these concepts
are considered and brought into relation with space and time we
have Dynamics and Mechanics, pure and applied. But the defi-
nitions and analysis so far given present simply the logical possi-
bility of an existential motion, rest, and impenetrability. They
show what characteristics an existential motion or rest must have,
although these may also have other characteristics. By them-
selves they show what motion and rest are, whether or not these
exist. For, granting that in order to have an existential motion or
rest, there must be something, say, matter, at rest or in motion,
then motion and rest are not this matter, just as they are not time
and space.

These propositions and the position with which they are identical
may seem to involve a degree of analysis which might well be open
to question and to offer a vulnerable opening for attack. But
even if this opportunity were improved and were successful, it

1 Therefore I have symbolized the individual R's by r', r", etc., in the above
formulae for motion.

198 DEFENSE OF ANALYSIS

would not invalidate our general defense of analysis against a
specific type of attack. It could show at best only that this
specific analysis and defense is in error, an outcome the possi-
bility of which the most strenuous supporter of analysis would
admit.

The supporter of the realistic view of the analysis of motion,
etc., can make his defense, then, even though, for example, motion
could not be defined unless there were something to move. For let
this supposition be granted. Let it be granted that there could
be no motion, or rest, unless there were a material particle or
particles. The fact is that motion and rest exist, for there are
material particles and bodies. And the actual attack on the analy-
sis of motion concerns the existential motion. The defense, there-
fore, can be made on the same basis, although it is not neces-
sarily limited to this.

The attack has been stated. It is argued that the analysis of
motion leads to rests, and so is false ; or, that it leads from con-
tinuity to discontinuity, and so, also, is false. This attack gains
plausibility, because, once again, there is a misstatement of the
actual results of analysis ; the character of the terms is misstated,
and the organizing relation is neglected. For consider an actual
case of motion, that of a body 0 moving from the point or position
A to the point or position B. Some means of measurement is
necessary to determine, within certain limits of accuracy, certain
quantities involved in this motion, namely, the distance traveled,
the time required for this, etc. But once this is done, the following
state of affairs is recognized by the analyst. The path, whatever
be its curve, is a continuous series of points, and the time a con-
tinuous series of instants. And the material particle, or its center
of gravity (centroid), serves existentially to correlate each specific
point of the path with one and only one specific instant of time.
For if there were a correlation of one and the same point with two
instants, the particle would be at rest. Put in terms of the pre-
ceding kinematical analysis, the material particle serves to make
existential a certain series of correlating one-one relations between

MOTION ANALYZED 199

the points of the space which is the path and the instants of time
required. And just as both the space series and the time series are
continuous and are in one-one correspondence with the real num-
bers, so is the series of specific individual correlating relations with
their terms continuous, with these complex terms related asym-
metrically and transitively and in one-one correspondence with
the real numbers. The existential motion is the material particle
making existential the series of one-one correlating relations be-
tween the points of space and the instants of time. It is all this.
Making this series existential is the motion of the particle.

It can be shown very clearly, now, that the results of this analysis,
which is really identical with the results obtained by applying the
calculus to motion, are not open to certain interpretations, namely,
just those on the basis of which the attack is made. Thus the one-one
correlation of specific instants with specific points, by means of the
material particle, that is, the correlation of one point y with one
and only one instant x and not with this and another instant xr,
is not to be interpreted as rest. It may be said to be an instance
of 'occupying/1 by one entity, of one point at a time, but this
term is somewhat vague and ambiguous, and needs definition.
Rest is undoubtedly in some sense different from motion. And
the only way to state this difference with precision is to define ex-
istential rest as involving the many-one correlation, by a material
particle, of many instants with one point. Therefore, if, in the
analysis of motion, any one of the manifold of complexes, each
consisting of a one-one relation correlating one instant with one
point, be interpreted as rest, the above definition is presupposed,
this one instant itself is implied to be, not one, but many, and
a contradiction is introduced. In every case here we are led
ultimately to terms not further analyzable into terms of the same
kind. Instants and points are such terms in the case of time
and space, and the complex term, xRy, under discussion here, is in
the case of motion. The occupying of a point at an instant by
one entity is, therefore, not to be interpreted as rest. It cannot

1 Mr. Russell's term, op. dt., 465.

200 DEFENSE OF ANALYSIS

be, without implying the contradictory, that it is not to be so
interpreted. The arbitrary attempt discloses that which is not
arbitrary, but is ultimate, necessary fact.

But just as little is this complex, this one-one correlation of a
point with an instant, the series of which complexes is motion, it-
self to be interpreted as motion. Sometimes it is supposed that
it must be so interpreted. It is supposed that motion presupposes
or implies other, partial motions, and these in turn other, smaller
motions — no matter how small — and so on. But this involves
or is identical with an infinite regress. Then motion cannot be
denned except in circular terms, or parts other than motions be
discovered, unless this infinite regress can be escaped from or
avoided. But escape is, by the very nature of the regress, im-
possible.

As we have previously seen, infinite wholes are of two kinds,
objectionable and unobjectionable, those involving difficulties,
and those not. Any infinite whole or series presents difficulties,
if the attempt is made to treat it only by enumeration. For
then, strictly speaking, it cannot be treated in its infinite character.

Indeed, to attempt to treat an infinite series by enumeration is
already to involve oneself in an infinite regress. For enumeration
serves to discover or to count only further terms like those already
enumerated, which further terms imply further similar terms to
be enumerated. An infinite regress, then, is never completed,
and it is impossible on its basis to get other than a circular defi-
nition. An infinite regress is for this reason, if for no other,
objectionable. Sometimes, however, we find the position taken
that the very nature of the infinite is its incompleteness or *un-
completedness.' Indeed this definition is sometimes employed in
order to attack realism. It is held that realism means, among other
things, the completed infinite — of space, for example. But then,
it is argued, the completed infinite is self-contradictory. There-
fore the conclusion is drawn that realism involves a specific self-
contradiction ! Manifestly, however, this argument is based on
the tacitly held premise that the infinite can be treated only by

INFINITE WHOLES 201

enumeration. Then, it follows that the infinite is never completely
treated, that is, is uncompleted. The definition of the infinite
which follows from the premise assumed illustrates with precision
the outcome of that premise, and confirms the initial assertion.
That definition is that the infinite is the uncompleted, the unended.
But this is only a circular definition etymologically obscure and
well illustrating the infinite regress which any circular definition
involves.

How different from the foregoing definition is the non-circular
definition that an infinite whole is such a whole as is similar to a
proper part of itself by virtue of both whole and proper part
being composed of terms which are different from both whole and
proper part — proper part being defined as a part which is similar
qualitatively to the whole. Thus, a finite line is similar to a proper
part of itself, in that, line and part being one-dimensional spatial
extension, both are composed of an infinite number of points in an
asymmetrical transitive relation. And the case is the same with
any finite period of time, and with any segment of the real num-
bers. But to deal with the infinite by means of such a definition
is to deal with it by implication, by intension, and not by enumera-
tion. Motion does presuppose smaller motions, a line, smaller
lines, but in each one of these cases there are also other entities
presupposed which are different from the whole, namely, the xRy's
(instant-R-point) in the case of motion, and points in the case
of lines, and these entities cannot be discovered by enumeration.

There is a kind of infinite whole, therefore, which is quite unob-
jectionable, which can be defined by a non-circular definition (thus
avoiding the regress), and which is both complete and completed
in just the way indicated by the definition; that is, it is dealt with
in such a way that we can discover something about it and some-
thing by means of it. Indeed, this infinite is the only one which
throws any light on continuity. This can be demonstrated as fol-
lows : The integers form an infinite series, but they are not con-
tinuous, for there is not one integer between any two. They can
be treated and defined either by enumeration or by intension.

202 DEFENSE OF ANALYSIS

By the one method they are an endless series. By the other,
they are such a series as implies, for example, that the class of
even integers has as many members as has the whole class, for
the even integers can be put in one-one correspondence with all
the integers. Therefore the integers are infinite, but they are not
continuous, and likewise with the rational fractions. Infinity
does not imply continuity, but the converse proposition does
not hold. Continuity does imply infinity, but not alone that in-
finity, which, such as that of the integers and rational fractions,
can be dealt with by enumeration, but also that infinity which,
as in the case of the irrationals, can be dealt with only by intension.
To get at the nature of continuity we must have, therefore, some
other method than that of enumeration, some method other than
one which gets us into an infinite regress. Motion does presuppose
motion, extension, smaller extensions.1 But to recognize this
gets us nowhere, helps us not at all in getting terms which are
different from motion and extension, and does not make possible
a non-circular definition. Quite similarly it would seem to be
a useless interpretation to define a point or an instant as either
extended or unextended, in contradiction to the extension of space
and time. The use of either the positive or negative term really
gives a circular definition. Extension and 'un-extension/ con-
tinuity and discontinuity, have nothing to do with points or in-
stants as such. Likewise the complex terms, the one-one relations
between specific instants of time and specific points of space in
the case of any actual specific motion, are neither rest nor motion,
neither continuous nor discontinuous, neither spatial nor temporal
(in isolation), nor extension. They are what they are, complexes
of existential one-one relations correlating the points of a specific
finite space and the instants of a specific period of time, the series
of such complex terms being motion.

I may now summarize my reply to this first attack on the
analysis of motion which claims that the analysis is false because
it leads to rests. Motion can be admitted to presuppose smaller
motions, but this gets us nowhere, and is identical with treating

ATTACK AND REPLY 203

an infinite whole only enumeratively — a method which makes
the infinite the uncompleted, and involves us in a number of artificial
antinomies and contradictions. On the other hand, if that method
is used which is the only possible method of actually dealing with
infinite wholes in their infinity, namely, the method of intension,
then motion is found to be a continuous series of one-one correlating
relations with their terms, these correlating relations and the com-
plex terms which they form being themselves related asymmetri-
cally and transitively. Motion is, then, the field of an asymmetri-
cal transitive relation whose terms are again relations that are
made existential by a material particle. The terms are not prima
fade the contradictory of the original entity analyzed, the motion.
They are simply different from it. Neglect the organizing relation
between them, and the terms will seem to be rests, and some
kind of 'transition' from term to term will seem to be necessary.
Consider this relation, and observe the true character of the analy-
sis, and the terms are found to be neither rest nor motion. So,
too, if the true character both of the terms and of the relation
between them is observed, a concise and exact definition and ex-
planation of the continuity of the original motion is obtainable,
in a manner quite analogous to the definition and explanation of
the continuity of space and of time.

The second argument against the analysis of motion, that it
leads from the original continuity to a discontinuity between the
terms, fails also. For here, as with the points of space and the
instants of time, the alleged discontinuity is that very charac-
teristic by which the continuity of a series is defined — the 'parti-
tion-definition' of Dedekind.1

It must be concluded, that the analysis of motion, as made by
modern science, stands unimpugned. It is adequate, and it is not
false. It reveals terms and relations which by themselves are dif-
ferent from the whole, but the additional characteristics of the whole
are also revealed. There is a creative synthesis ; the whole as a
whole is different from the terms and relations taken individually,

1 This section, on Space.

204 DEFENSE OF ANALYSIS

but it is these parts related. Then the analysis is adequate. And
it is not false, since it is not found to lead, when stated correctly, to
terms or relations which in any way are the contradictories of the
original whole. The terms are not rests, they are not motions, they
are not discontinuous from one another, and there is no transition
from one to another. They are, as a series, translation, motion.
But with the analysis qua analysis thus vindicated and with the
general realistic position assumed to be established, it may be said
that both the motion as a whole, and the transitive asymmetrical
relations, the correlating relations, and the points and instants
involved in that whole, are all equally real, although not the same
kind of realities. The analysis of motion, like the analysis of space
and time, is a method of discovering entities which are independent
of their discovery and of their being known.

5. Velocity and Acceleration. — Just as the one-one correlating
relations, themselves related asymmetrically and transitively
(which with their terms, instants, and points, constitute motion),
are existing relations, if the correlation is effected by a material
particle, so also are velocity and acceleration existing enti-
ties under the same condition. In fact, there is no actual
motion that does not involve velocity, or a negative or posi-
tive acceleration. But these characteristics are not involved
necessarily in motion as such, even as motion is not involved in
space or in time as such, or these in each other, or both together in
logical principles. Analysis reveals a most interesting and
peculiar relation, that of logical priority.1 This relation, as an
analytical result, is, of course, attacked, but it can be successfully
defended. It involves nothing contradictory if it is treated on
a strictly empirical basis.2 Logical priority may be defined as
that state of affairs in which, for example, a proposition B implies
or presupposes proposition A, but A does not imply B. A is

1 Cf. Marvin and Perry, this volume, also Russell, op. cit., 114.
* Speculatively, of course, it can be attacked. But such an attack is replied to
in Section IV of this essay.

205

then logically prior to B. The instances of this relation are
many, and the fact of its subsistence is evidence for a pluralistic
universe. Certain A's could exist or subsist without B's, al-
though, conversely, not. Certain important instances of this
relation concern us in the present discussion. Thus, a propos
of the statement just made concerning motion, etc., it may be
said that logical principles are logically prior to all else, — to
space, to time, etc., — such logical principles, namely, as in-
clude the principles of pure mathematics (arithmetic). Space
and time seem to be coordinate, but they are logically prior to
motion as such. Change, however, is logically prior to motion,
and motion is logically prior to its causation and to its constant
or changing velocity.

The further examination of the field of this relation of logical
priority shows, moreover, that there are certain sciences which are
logically prior to others. Thus, letting the order of enumeration
stand for the relation under discussion, some of the sciences may
be arranged in the following series in order of logical priority:
Logic, Geometry, Science of Time, Kinematics, Dynamics, Pure
Mechanics, Applied Mechanics, Physics, Physical Chemistry,
Physiology, Psychology.

The logic here meant includes not only the logic of propositions,
of classification, of definitions, as usually understood in the text-
books, but also the logic of relations, of infinite classes, of con-
tinuity, of variable and constant, of the status of entities, etc., etc.,
some of which has been made use of in this essay. Some of the
important features of geometry and of the science of time have
been presented, and the demonstration made that space and
time imply certain logical principles, although the converse
is not the case; and the analysis of motion as such that has
been presented constitutes part of Kinematics, implied by Dy-
namics, but not conversely. But we are now to consider, briefly,
velocity and acceleration. That will take us beyond Kinematics.
Motion as such has been analyzed and defined without reference
to these entities, but, very evidently, they imply motion. And

206 DEFENSE OF ANALYSIS

existentially, of course, there is no motion which does not involve
velocity — either constant or changing. What, now, are velocity
and acceleration ? Once again, as in the discovery of actual cases
of motion and in the statement of what it is, whether discovered
or not, a distinction must be made between the means of discover-
ing velocity and acceleration and these entities themselves. To
discover actual motion there is necessary the empirical observation,
either direct or indirect, of two points, A and B, of one configuration
defined with reference to one set of coordinates, and the observation
of the traveling of a body from A to B along some path. To dis-
cover velocity, not only are these two observations necessary,
but also the empirical observation and measurement of the time
taken for the body to move from A to J5, and of the length of the
path AB. We then have two finite quantities, distance traveled,
and time taken, and a material body. The motion of the body is,
then, that whole which is the series of relations correlating the
points of this path with the instants of this time. The relation
of the distance traveled and the time taken can then be expressed,

o

assuming the measurement to be accurate, in fractional form, -,

I

and this is velocity as commonly understood. It has a definite
numerical value, found through the measurement of the finite
time and space involved. So far, then, velocity is an entity, an
existential relation between a specific time and space, mediated
by a material particle. Although expressed by a fraction, it is
one entity, though perhaps a complex one, and not two, for a frac-
tion, although symbolized by two or more signs, is only one number.
But the velocity, or rather its value, is so far only an average ve-
locity, for the same resultant would be obtained if during the time
taken for the whole motion there were plus and minus deviations
from this average. The velocity would be presumed to be con-
stant provided only, that, if found for any distance As, no matter
how small, for any time A£, no matter how small, it would re-
quire the time nA£ to travel the distance nAs, where n is the same
multiplier. But this would be only a presumption ; any velocity,

VELOCITY 207

ds

— , might be only an average of deviating velocities. For practical

(K

purposes, of course, average velocities are quite sufficient. Yet
modern analysis is theoretically able to determine velocity for any
instant of a body's finite motion, and so to show that there are
deviations, or that the velocity is genuinely constant, that is, the
same at every instant. The velocity can be shown to be constant,
or not, with a greater degree of assurance, by taking any two s's
traveled in two t's, passing to the limit of each of the two ratios,
and comparing the -values thus obtained. In fact, only in this
way can it be determined whether the velocity is constant or not.
One determination for any one As and A£, whether this be part
or whole, relatively small or large, does not suffice to do this, but
shows only what the velocity is at a certain instant or what its
average is for a certain finite space or time. Two determinations,
either of the whole and of any partial motion, or of any two partial
motions, alone suffice to decide the question, under the above con-
ditions, whether the velocity is constant or not.

But with this question once decided, what are the velocity
and the acceleration ? Let us consider first the case in which the
velocity is constant. In this case the numerical value of the limit

As
of any and of all the ratios, — , is the same for every instant of the

time required for the motion. This value is the value of the complex,
the one-one relation correlating each point of the path with each

As
instant of the time taken. For — taken at the limit means, as

As

does the ratio itself, space-related-to-time. For — , it is finite dis-
tance related to finite time ; for the limit, it is point related to
instant. In the case of a motion with constant velocity, therefore,
the complexes, the correlating relations with their terms, have
values ; they are magnitudes, since they can be greater or less than,
or equal to something else of the same kind, that is, some other
velocity ; but as constant velocities of one motion they have the

208 DEFENSE OF ANALYSIS

same value, or are the same magnitude, expressed by the same
number — whatever this may be. How, then, can they be
the terms of a series, the field of an asymmetrical transitive
relation — which they must be as constituting motion? For
does not this relation demand different values, values in
order of magnitude 1 The answer to the question is that the
latter point concerns only the continuity of the motion, and not
the constancy of the velocity. The continuity of motion is the
same as the continuity of the real numbers, the number con-
tinuum. There is a one-one correspondence between these
numbers and the correlating relations which with their terms are
motion. Then these complexes have an order. But at the same
time, each complex, consisting of correlating relation with its terms,
has a value or is a magnitude, and this value in the case of constant
velocity is the same for all these complex terms. That value may
be expressed by any real number, but, once found for constant
velocity, it is the value of the complex, correlating relation with
its terms, for every instant.

I must disagree, then, with the statement that "there is no
such thing as velocity except in the sense of a real number which
is the limit for a certain set of ratios." l Existential velocity is the
magnitude of the existential complex, consisting of the one-one
relation between one instant and one point, when this relation is
mediated by a material particle. It itself is not time, nor space,
nor matter; but it is involved hi these; that is, it exists if there is a
real material particle moving and so serving to make the continuous
series of correlating relations existential. But there are moving
material particles. Then velocity exists, although it is a complex,
that is, a relation and its terms. And, existing, it is also a magni-
tude in that it is equal to or greater or less than other velocities.
It is, therefore, both a complex and a magnitude. Once dis-
covered by taking the ratio of small distances traveled to small
times required, and passing to the limit, it can, however, be defined
independently of this method, even as has been done above. But

» Russell, op. tit., 473.

ACCELERATION 209

in turn, velocity as such can be defined independently of existen-
tial velocities. The latter exist if material particles exist and
move. But they would subsist in that the complex referred to is a
magnitude. Thus there are possible, subsisting velocities which
have not existed, do not exist, and perhaps never will. There
is, then, a kinematical definition of velocity as well as a dynamical
one.

ACCELEEATION

The case with acceleration is much the same as with velocity.
Once discovered, it can be defined independently of that discovery,
and even independently of existing accelerations. The method for
ascertaining whether or not there is acceleration, positive or nega-
tive, is that of finding out whether there is constant velocity. Two

ds

-r's must be obtained on the basis of empirical measurement, and

compared; that is, the numerical values of the velocities at two
instants must be determined and compared. Let these values
be different. Then they are but different values of the magnitude
of the complex, correlating relation between a specific point and
instant. They exist if the correlating relation is mediated by a
material particle. Its mediation of all the correlating relations
in a certain finite series is its motion. But if the material particle
moves, then there may be a real acceleration. To determine,
in turn, whether the acceleration is uniform or not, demands
methods which, since they are quite analogous to those required
for determining whether velocity is uniform or not,1 need not be
presented here. But let it be found that the acceleration in a
number of specific instances is uniform. Then the generalization
can be stated, and a definition of uniform acceleration be given which
is independent of the method of discovery. Uniform acceleration
then turns out to be a most interesting entity. Motion is change
of position. It is a whole, a series, as has been explained, with its
terms complexes which are neither rest nor motion. Velocity is

1 See this section, on Motion.

P

210 DEFENSE OF ANALYSIS

a correlating relation and its terms, the whole having magnitude.
It is identical with any one of the terms of motion in respect to
their magnitude. Acceleration, now, is a whole, a series, a change.
It is the change of velocity in time, the second derivative. It in-
volves, then, the one-one correlation of the terms of a series of ve-
locities with the instants of time. We have, therefore, a complex
term, pR\i (= velocity, the correlation of a point p with an in-
stant i) in turn in one-one correlation with an instant, thus, —
(pRii) Rzi or vR2i. Put in symbolic form, and using r = R2, r
being the correlating relation, as in the formula for motion1, we
have

v v' v" v"r

rRr'Rr"Rrm - • -

i i' i" i'"

Acceleration is, then, the series of complex terms, each one a
(pRii) Rzi, related asymmetrically and transitively. If a material
particle mediates this series of correlating relations which is mo-
tion, and the velocity changes, then the acceleration is existential.
Yet it can be defined independently of the supposition that it is so
mediated, as the above discussion makes evident.

Acceleration is, therefore, itself change ; it is in this respect, like
motion, a whole. Analysis of it shows that it, too, like motion,
is composed of complex terms which are neither rest nor motion.
They are just what they are — complex terms consisting of a one-
one relation correlating each velocity with an instant, each pRii
with an i. But these terms form a series corresponding to the
time series with which they are correlated. That series is, then,
continuous, and its terms are infinite. Specific accelerations can,
like velocities, be compared and their difference or equality ascer-
tained. All accelerations are, therefore, magnitudes, and so are in
one-one correspondence with the real numbers, for all accelerations
are possible, though not existential. A specific uniform accelera-
tion is, then, a continuous series of individual velocities, but of

> See the previous section.

ATTACK AND REPLY 211

all velocities between certain limits, and these velocities would
ipso facto have, between such limits, the natural order of the real
numbers with which they are in one-one correspondence. If the
acceleration is uniform, then, if there is an acceleration of finite
amount a in the time b, in the multiple time nb there will be a
change of velocity of amount na.1

With reference, now, to the attack which might be made on
the analysis both of velocity and of acceleration along the same
lines that it is brought against the analysis of motion, it is evident
that it is invalidated in the former case for the same reasons that
it is in the latter. Velocity is the complex term composed of
the correlating relation between a specific point and a specific
instant, together with the magnitude of this relation. It is, then,
neither rest nor motion, so there is no opportunity for the attack
here. And just as motion is this continuous series of complex
terms, themselves related asymmetrically and transitively, so is
the constancy of velocity only the constancy of the magnitude of
all the terms of this series. The constancy is a continuity of magni-
tude. The terms together with the asymmetrical, transitive rela-
tions define this continuity with precision and adequacy. Thus de-
fined, there is no more opportunity for introducing a discontinuity
of terms or of 'partitions/ and so of claiming contradiction and
falsification, than there is in the case of motion, where the oppor-
tunity is found to be negative.

The case with acceleration is similar. Acceleration is change
of velocity. Uniform acceleration is continuous change of velocity,
but in neither case are the terms either change or rest. They
are complexes, velocities in one-one correlation with specific in-
stants of time. But no one such complex individual is either
change or rest. The change of velocity, the acceleration, is the
series of such terms related asymmetrically and transitively, and
involves a one-one correlation with the instants of the time series.
Absence of acceleration, or what might be called 'resting velocity,'
is analytically a many-one correlation of one velocity with many

1 Cf . this section, on Numbers.

212 DEFENSE OF ANALYSIS

instants. In no way is there opportunity here for the typical
attack. Further, the usual typical claim that this analysis in-
troduces that which is discontinuous into the continuous, in that it
makes the terms discrete or allows of a 'partition,' is but another
way of stating the very continuity of uniform acceleration.1 The
attack fails here also, and the analysis is to be accepted as precise
and adequate, and is to be interpreted realistically. Acceleration,
like velocity and motion, is not a mere number.2 It is a series — of
complex terms ; it exists if a material particle moves with a velocity
which is not constant, otherwise it subsists. All accelerations,
like all velocities, are logically possible, though not all are found in
or implied by our existential world, as existing or even as possibly
existing.

DYNAMICS AND DURATION

It has been shown in the above discussion of the analysis of mo-
tion, of velocity and of acceleration, that these entities, once dis-
covered, can be defined in a purely logical way without introducing
existential conditions into the definitions. The science which
treats of these entities as so defined is Kinematics. Such entities
do, of course, exist in an extremely large number of specific in-
stances, but the treatment of them concerns not only the existen-
tial, but also the subsistential cases. Kinematics is logically prior
to the sciences of existential motion, velocities, and accelerations.

The preceding discussion puts us in a position to advance to the
consideration of what is the next logical step in the grouping of
the sciences under examination, and to get a basis for the refuta-
tion of still another attack on analysis, namely, that attack which
Bergson presents in his doctrine of duration.3

Motion, velocity, and acceleration have been discussed. Thus
far, however, there has been no discussion of causation. Yet exis-
tentially, in the specific cases where these entities are discovered,

1 See this section, on Motion and on Space.

1 Cf. Russell's dissenting statement, Principles of Mathematics, 473.

* Creative Evolution and Matter and Memory in numerous places.

DYNAMICS AND DURATION 213

causes are found to operate — causes for motion, for a change of
velocity or of direction, etc., etc., — in fact, causes in general.
Here, however, I am not concerned with all causes or even with
the principle of causation, but only with one type, that, namely,
which is made the pivot in the specific attack which I wish to
confute. That type of cause is that one in virtue of which an effect
at any time whatsoever, past, present, or future, near or remote, can
be inferred or discovered. The position taken by the attacking
party here is, that this means that the whole temporal series of
events, that is, the kind of temporal series to which analysis leads,
is given all at once, now, but that this contradicts the very nature
of time as lived, as experienced, as related to, and so as constituting
things in time. As lived or experienced, etc., the attack continues,
tune, or 'duration,' as it is called, is not given all at once — now;
but life and experience are immersed in it, and it is immersed in them,
in such a way that it is only a now and a now and a now that is actually
lived, although each now is an accumulative effect of all that has
gone before. Since the time with which analysis deals is held to
contradict in this way the time as lived, and the latter is held to be
reality, the attacking party concludes that analysis falsifies the
original nature of tune. The argument is not different in form,
then, from the other specific attacks which we have considered.
Thus it attacks the realistic interpretation of tune which discovers
in it an entity that is absolute and not relative and that is genuinely
independent of existents. I shall show, however, that this spe-
cific attack presupposes this very realistic interpretation and so
contradicts itself. To that task I now set myself.

First, let us consider in some further detail the position which
is attacked. That position is one which introduces into, or adds
to the analysis which we have so far considered, the concept of
causation and even of change in general. The attack is directed,
then, against that body of analysis which is called Dynamics.
Or, indeed, further than this, the attack is made also on that body
of analysis which introduces still more specific causes and con-
ditions than does Dynamics, namely, Mechanics.

214 DEFENSE OF ANALYSIS

Motion may be vaguely said to be change of position — in time,
of course. But this change of position can be accurately and
adequately analyzed. There are, however, other kinds of changes,
as, for example, change of velocity and qualitative changes. In
fact change in general is not only genus to any specific type of
change, but is logically prior to it. Any specific type of change,
such as chemical reactions, electrical and thermic events, involves
time, and can be referred to it as the independent variable.1 Put
in familiar scientific terms, this means that, with any qualitative
change measured, and the time in which it takes place also meas-
ured, the average velocity of the change can be determined, and
the whole process treated in perfect analogy to the treatment
of motion. Thus, the average velocity of a chemical reaction

is the limit, ^, of the ratio amount of new substance produced,
dt time taken for this

that is, a;i~ . To make certain that this is not simply an average,
tz — t\

but is a constant velocity holding good for any instant of time,
or to ascertain that there is a uniform acceleration here,
demands the same observations, etc., as for motion.2 Once ascer-
tained, however, the change, the velocity, or the acceleration can
be defined without introducing these conditions into the definition.
Further, and important, exactly the same interpretation of all
these 'non-motion' changes must be made as is made of change of
position. The terms are neither rests nor changes. They are com-
plex terms of one-one correlating relations between the instants of
time and some qualitative entity (which may in turn be complex)
— with these complex terms themselves related asymmetrically
and transitively. The whole series, with all its terms and rela-
tions, is the change.

In customary scientific language any finite change x is, then, a
function of a certain specific time t as the independent variable,
and can be graphically represented by the ' time chart ' :

1 See this section, on Relations.

* See the three previous divisions of this section.

TIME CHART

215

0

dx
Not only can the first derivative, — , that is, the velocity of any

Civ

d?x

change, be thus represented, but also the second derivative, — ,

ctt

the change of this velocity, and, theoretically, the higher deriva-

dnx

tives, - — , although those beyond the second are seldom used.
dtn

All this is important as leading up to the presentation of the
meaning of a dynamical system and of making clear the position
which is attacked by the doctrine of ' duration.'

The 'time chart' makes use of only two coordinates, and these
are all that it needs. But to determine one of these coordinates,
x, three or more coordinates may be necessary. Thus, to determine
the extent of any specific motion, three coordinates forming one
' frame of reference ' are necessary, although, once determined, this
extent can be represented on one coordinate, referring the motion,
with either its constant velocity or acceleration, to time. Suppose

216 DEFENSE OF ANALYSIS

such a determination to be made for any specific change whatso-
ever, for example, for any qualitative change, or for the change of
velocity of any such change; that is, let the amount of the change
be determined and represented on the 'time chart.' Then, on the
assumption that the objective change, represented by the chart,
is continuous — and this is the supposition at this point in our
discussion — the state of affairs at any instant k, before, after,
or between the instants t and ti is determinable, and the objective
change is, of course, determined or caused. This gives a view of
causation which differs quite radically from the traditional view.1
Briefly, it means that in a continuous change of any kind, imply-
ing the one-one correlation between the instants of time and the
points of space, or of that which, as specified by coordinates, is the
equivalent of a point, namely, a configuration, any five terms of
the group of terms, C, C\, C2, t, ti, &, determine a sixth. Pre-
sented as a formula this means that C = F (Ci, t\, Cz, U, O-2

All this may now be made somewhat clearer. First, a dis-
tinction must be made between the cognitive determination or
specification of the value of the terms concerned, and the causal
determination. Thus, hi the above formula, if the terms to the
right are known, C can be ascertained. But the knowledge will
be correct, provided only there is a genuine objective causal de-
termination, of the kind above defined, between the configurations
at any two times and that at a third time. It is the objective
thing that we wish to consider now, distinguishing it from the
knowledge of it. Making this distinction, and adhering to it, the
statement made just above concerning causation can be expanded
as follows :

1. In the case of a material particle moving with constant ve-
locity, the velocity at any instant is determined or caused by the
velocity at any other two instants.

2. In the case of the uniform acceleration of the motion of a
body, both the motion and the acceleration for any finite tune be-

» See Russell, op. tit., LIV, LV, LVI, LVII et passim.
1 See Russell, ibid., 486.

CAUSATION 217

tween any two instants, t\— k, is determined by the acceleration for
any other finite time, tx— tv, before, after, or within h— U, where x
and y are definite and specific.1

3. Exactly the same statement holds of changes which are not
motions, and of their velocities and accelerations; or, since ac-
celeration is change of velocity, the statements hold good of
change in general, provided it is continuous. If it be continuous,
no matter what its nature may be, the principle of causation which
we are discussing may be expressed in the above formula.

The simplest existential case of such causation is the motion of
a material particle, but all sorts of complex cases are also possible,
and the same principle holds for them. For, no matter how com-
plex and varied the factors in a state of affairs may be, they can
be specified by the use of coordinates, and these coordinates de-
termine what is logically equivalent to a point. That a complex
state of affairs can be so specified presupposes that it is objectively
specific in some such sense as the coordinates indicate.

Now there are actually a great many other factors involved in
the objective specification of a system of the kind we are consid-
ering; but we need not examine these in any detail in order to make
our point against the attack to which we are replying. Suffice
it to say, that a very complex system can in due time, by means of
the composition of vectors, the Hamiltonian principle, etc., be
specified by three coordinates.2 A change in these coordinates,
representing a change of any kind, can be represented subsequently
on the time chart as if it were a change of position.

With the distinction between the object known and the knowing
granted, and with it also granted that there is an objective causal
determination of the kind just defined, in reality complex, yet
capable of a simplifying treatment, it follows that, if the numerical
value of the configurations at two times be known, the configura-
tion at any other time, be it ever so complex, can also become

1 The acceleration is determinable, i.e., known, however, only if x and y are known:
then three quantities or values will be determinable.

1 Cf., e.g., Webster, L. G., The Dynamics of a Particle, 1908.

218 DEFENSE OF ANALYSIS

known — if, as is assumed, the change is continuous. It is by
dealing with things in this manner, therefore, that time does seem to be
defied in some way, even as the attacking party claims, and that
at the present moment, as it were, we can reach out indefinitely
into both past and future, so that everything is given now. And
this does seem to furnish a marked contrast with experiences in our
own lives which we must wait upon with infinite patience and self-
control, or which are irrevocable. But there is a serious ambiguity
in the contrast, and so, also, in the conclusion drawn from it, —
for it is on the basis of this contrast that the specific attack under
consideration is made. The analysis which leads to this bringing
of the past and the future into the very present, and which in this
sense seems to do away with all real time distinctions, can only be
a falsification, the attacking party holds, of real time or duration.
We shall shortly demonstrate, however, that this attack involves
much confusion and is quite easily refuted.

That analysis, however, which is thus accused of foreshortening
both past and future to the confines of the present, contains
many elements, and these, in accordance with the previously pre-
sented series of sciences, may now be distinguished to advantage.

The science of numbers, of tune, of space, and of motion have
each been considered in detail. Motion can be defined, it has been
found, in a purely logical manner, and the science of motion thus
defined is Kinematics. Kinematics is applied when there are
actual particles of matter moving, but pure Kinematics is logi-
cally prior to such an application. In Dynamics, to distinguish
it from Kinematics, causation and configurations of entities are
taken into consideration. The configurations might in pure
Dynamics be limited to positions and motions, or to some kind of
equilibrium or change, or to any combination of these, and then,
by specifying any two of these at the times t and t\, a third at U
would be specifiable, if the change were continuous.

But in applied Dynamics the configurations would only be of
a certain type, each configuration, specifiable by coordinates,
being correlated with a specific instant of time, itself specified in

KINEMATICS, DYNAMICS AND MECHANICS 219

relation to our actual existent world; that is, the configurations
would be only those which we find empirically in our actual world.
But here the changes are not continuous, or, rather, not unlimited,
although they are continuous between certain limits, so that, were
the inference made that at a certain time it, beyond a certain finite
period between t and t\, a certain configuration would be deter-
minate and discoverable, this inference would be in error.1 The exis-
tential world does not seem in respect to all of its attributes to be
of a uniformly continuous character. Rather, there seem to be dis-
continuity, critical points, both in the temporal series of natural and
artificial processes and in formal synthesis. Thus, for example, an
organism presents at certain stages of its development seemingly
quite new characters, a physico-chemical substance goes through
critical points in passing from the gaseous to the liquid and solid
states, and the synthesis of chemical compounds out of components
leads to the appearance of new properties.

Pure Dynamics, however, finds no place for these discontinuities.
It introduces the concept of cause, and adheres to the principle of
continuity. It can construct any number of dynamical systems;
but only one of these accords with the actual world, and this is
one which allows of the application of pure Dynamics only within
certain specific ranges whose limits are empirically determined.

Thus, with Dynamics introducing the concept of causation, the
next step in the series of sciences is the specification of the types
of causation. If these are made the attraction and repulsion of
material bodies, — attraction in the inverse square of the distance,
repulsion in accordance with the 'laws of motion/ we then have
the Newtonian Mechanics. Make the law of attraction in some
other ratio, that is, allow for any ratio, and we have a general-
ized, a pure Mechanics. But the law of inverse squares, for ex-
ample, is found empirically not to hold for all distances; thus, it
does not apply to intramolecular distances. Such limitations
bring us to the realm of Physics. With certain limiting points, and,
therefore, ranges of qualitatively different phenomena empirically

1 Cf. Montague, this volume.

220 DEFENSE OF ANALYSIS

determined, Arithmetic, Geometry, Kinematics, Dynamics, and
Mechanics are applied, and are presupposed, each by each, in the
inverse of the order named. But, conversely, each one of these
does not imply the propositions of the subsequently named science.
Each of these sciences is a science of a certain stratum of reality,
as it were. Each stratum has its peculiar characteristics which
can be discovered only empirically, and cannot be deduced from
the characteristics of the preceding stratum, but each of the logi-
cally, and, possibly, also temporally subsequent strata is com-
patible with the existence, or subsistence of the prior strata, and
so allows of the application to its phenomena of the sciences
which treat of the prior phenomena.

In this way we might continue, and consider the character of
Chemistry, of Physiology, etc., in their relation to the sciences
which we have considered, but for our purposes we do not need to
do this. Suffice it to say of the relation of these sciences to the ones
already discussed, that they presuppose the facts which these
other sciences treat of, but are not implied by them. Each science
in the order given is presupposed by the facts of the subsequent
ones, but their facts are not implied by its facts. Certain of the
facts of each science are peculiar to it, and can be discovered only
empirically. But once discovered, not only are these facts found
to presuppose other facts, and so to demand the application of the
preceding sciences, but they set limits to this application. Specific
laws hold good only within certain ranges. Phenomena within
these ranges are arithmetical, spatial, temporal ; they sometimes
are changes with velocities and accelerations ; they are caused, and
they are continuous. But the continuity is limited by the limits
of the range. At these limits there is discontinuity, between
them continuity. This seems to be the actual status of the exis-
tential world. Phenomena exist or take place in different strata.
Laws are limited in the range of their validity, because the con-
tinuity of phenomena of which they are laws is limited. There
is, then, an existential pluralism. Certain phenomena presuppose
others, which in turn do not imply them, but can be conceived

DISCONTINUITIES 221

and discovered as lacking them. The former limit the range of
the validity or application of the laws of the latter, the latter
apply within a certain range, because they are presupposed.
Thus we get a logical or an ontological pluralism, with different
strata of entities, — numbers, points, instants, logical motion,
material particles, causes, specific kinds of causes, etc.

The fact, however, that the actual world of physical, chemical,
and physiological fact is neither purely kinematical, nor dynami-
cal, nor mechanical, etc., but is discontinuous at certain points,
does not do away with the possibility of the attack which we are
considering. For the discontinuities can be bridged empirically, —
this is part of the task of analysis, — and the relation between two
discontinuous ranges of phenomena ascertained. Then, on the sup-
position that this functional relation always holds good under the
same conditions, the claim can be made once more that past and
present and future, to an all-inclusive extent, are given now, at the
present time. For example, the vapor pressure of a solid increases
at a continuous rate with rising temperature, until, at its melting
point, the pressure suddenly increases at a new rate. This is the
critical point, and is constant if the external pressure compensating
the internal is constant, though it varies as the latter varies. The

dj)

derivative, ~, has two values at this critical point. The ratio between
dv

these values remains constant, or changes continuously, according
as the external pressure remains constant, or changes continuously.
Or, at least, the ratio is assumed so to remain. On this basis, not-
withstanding the discontinuities in nature, we can reach out into
the past and future and have them given now, quite as well as if
all nature were continuous without limit. We have only to say,
Continuity up to a certain point, discontinuity at that point,
continuity beyond it, and the ratio remaining the same, or itself
changing continuously, in order to bring a discontinuous field of
future or past phenomena within our present ken as readily as we
could an all-inclusive continuous field. Only, if there are these
existential discontinuities, they cannot be discovered deductively,

222 DEFENSE OF ANALYSIS

but only inductively, which means that there are existential rea-
sons which necessitate an inductive procedure.

Reply can now be made to the specific attack on the analysis of
time, and to the claim that duration is the real entity. Through
analysis future and past do seem to be given now — in some sense.
That must be admitted. In fact, both attack and defense do admit
it. But the attack claims that for this reason the very nature of
time (as duration) is vitiated and contradicted, in fact that it is
thus 'detemporalized' and falsified ! For, it is held, time as lived,
etc., that is, as duration, cannot be so manipulated. It, real time,
demands that we wait with patience for the occurrence of that
which we desire. Time manipulates us, not conversely. In just
this sense it forms an integral part of the actual course of events,
to be separated from them only at the cost of falsifying its real
nature.

Now in this attack the validity of the specific analysis of time
into instants is not directly impugned, or, at most, it is at-
tacked only subsequently to the specific attack now under con-
sideration. I shall show, however, that that very analysis is
presupposed by this latter attack. Indeed, that there is a certain
difference between time as analyzed out and time as lived must, I
think, be admitted, though it may not be a difference in time qua
time, but only one between specific, individual periods of time.
The attacking party admits this difference in the distinction which
he grants and emphasizes between processes which, he insists, hav-
ing taken place in the past, are organically incorporated or sum-
marized in the present, or which mil take place, and those which
are taking place and being lived now. But this makes it evident
that he finds no difficulty in somehow getting into some parts or
periods of a time series which are different from, or other than the
present, and clearly, he must grant the same privilege to the defend-
ing party. In fact this distinction is implied in identifying the
present as the present. But this distinction, which the attack-
ing party insists upon, is only the distinction between perceptual
time and conceptual. He himself grants both of these kinds of

DURATION EXAMINED 223

time in making his attack, but argues and concludes that the
former alone is real, the latter falsification.

In reply to this I shall show that that very characteristic which
is most emphasized in this attack, namely, that which, put con-
cretely, is called the living now, and, put abstractly, is the 'unma-
nipulatableness ' of time, really demands the validity of the analysis
of time out of the complex of things in time, and, further, the analy-
sis into instants, and the realistic view of time. Stated briefly, this
can be done by showing that it is only the realistic view of time
and this twofold analysis that account for the very characteris-
tics and attributes which the attacking party is emphasizing. For,
assume the hypothesis (to see how it works), that time is an entity
which subsists independently of things in time, and that accord-
ingly it can be analyzed out of the complex, ' things in time ' ; then
it follows that there is a very interesting and important difference
between the whole time series as treated conceptually, and certain
time periods as experienced now, — call these duration. For, by
this assumption, this very difference, this very characteristic which
is so emphasized in the living now, this compulsion on us to be
patient and wait for things to develop s ep by step, arises from
the one-one correlation of specific events with those specific instants or
periods (in the larger time series} which are the now. The now is just
this specific time period, consisting of these, and not those specific in-
dividual instants in a larger time series, and in this sense cannot be
manipulated. This is equivalent to saying that the same hypothesis
accounts for the distinction between perceived time and conceptual
time — a distinction which the attacking party tacitly insists upon
and uses. But the time as time in the two cases is not different. It
is only the periods or the individual parts that are different. For
analysis shows time to be a series, its terms to be instants, its re-
lating relations to be asymmetrical and transitive. It is absolute
and not relative. It subsists whether anything exists in it or not,
and whether it is known or not. This is the realistic view of time.
Assume this hypothesis, and then, if specific existential events take
place, they take place in correlation with specific individual in-

224 DEFENSE OF ANALYSIS

slants constituting definite individual periods of time. This cor-
relation is fixed, it cannot be altered. If the events are those of a
conscious organism, then they are to be lived now with waiting,
with patience, preceding their arrival at a fixed period of the time
series. If they are now past events, they are also fixed in the
time series, with a definite temporal distance and a causal relation
between them and existential events now, and so, also, if they
are future events. The assumption accounts for the very things
most emphasized by the attacking party. But knowing can reach
out to this past and future, can, within certain empirical limits of
accuracy, get at the many events in them in the temporal order,
while the living, the perceiving, as an event, is limited to the now.
The two are different, the distinction is justified. But it is ac-
counted for, and along with it the Bergsonian doctrine of duration,
only by the hypothesis of an independent time series, in fact, by
the whole realistic view of time. Time is an independent series.
It can be analyzed out, because it is already out. It is not an in-
tegral part of a complex which cannot be analyzed without falsi-
fication. Things are in time, but it is not in them except in the
sense that existents are correlated with specific instants. Their
temporal position is fixed as a configuration C correlated with a
present time t, and in a many-one relation with this t and two con-
figurations Ci and C2 correlated with ti and k, past or future. This
is really an analytical statement of Bergson's doctrine of duration.
The latter presupposes for its own explanation the very view of
time which it purposes to attack — presupposes the validity of the
very distinction which it itself is compelled to make.

I conclude, then, that that attack on the analysis of time which
claims that the time analyzed out is detemporalized, spatialized,
falsified, etc., is itself false. The discovery of time as an entity
which is independent of things in it and of its analysis into instants
stands as valid. In fact, only this realistic interpretation both of
time as a whole and of its analysis accounts for those very features
which are emphasized in this specific attack against such a dis-
covery and analysis.

ATOMS 225

6. Other Classes of Individuals, Atoms, etc. — My general scheme
now brings me to the consideration of a new type of whole lying
within this class of classes which are composed of individuals, a
type, namely, whose analysis to a large extent is experimental,
although not exclusively so. This type of whole is well illustrated
by any finite quantity of any pure chemical substance. Such an
actually existent whole in any specific instance is not identical
with the class-concept which denotes the class of which it is a
member. The concept iron is not iron, the concept mercury is
not mercury, any more than the class-concept number is a number.
The defense of the analysis of wholes of this type constitutes,
among other things, a defense of the atomic theory and of those
theories which, like the molecular and electron, are allied with
this.1

I need not relate, indeed, not even classify all the experiments
which can be made in the laboratory, and all the additional evi-
dence, experimental and other, which shows that a chemical sub-
stance consists of parts called atoms. I shall consider such ex-
periments and evidence only so far as is necessary in order to
make a defense.

The proposition that each pure chemical substance consists of
atoms is a conclusion which is reached through a hypothetical
syllogism of the general form : If p alone implies q, and q is, then
p is ; that is, if p is the only hypothesis which explains q or. a num-
ber of q's, and these exist, then p exists.2 This seems to con-
trovert the usual rule for the hypothetical syllogism according to
which the affirmation of the subsequent does not imply the affirm-
ation of the antecedent ; but in reality it does not. The assertion
of the subsequent carries with it the assertion of the antecedent
provided an unequivocal connection between antecedent and con-
sequent is established whereby it is shown that, although many
p's might explain one q, only one p can account for may q's. The

1 This essay, Section V.

1 Cf. Marvin, W. T., The Existential Proposition, Jour, of Phil., Psychol, etc.,
1911, 8, 447-491.
Q

226 DEFENSE OF ANALYSIS

conclusion that p exists, if it explains existing g's, is valid, if other
hypotheses are excluded through the fact that one and only one p
accounts in the present state of our knowledge, for all the known
facts, for all the q's. This is the type of reasoning which we have,
now, in the establishment of the atomic theory. Experimental
analysis reveals many q's, many existential facts. The only hy-
pothesis which explains them is the atomic, which means an analy-
sis of the whole into parts called atoms.

For the sake of making my presentation specific, I shall state
some of the important facts l which the analysis into atoms ex-
plains :

(a) Pure substances combine in constant mass-proportions, no
matter how small the quantities are which are worked with ex-
perimentally. This would be explained if larger masses are only
multiples of some ultimate units between which the same mass-
proportions hold.

(6) Certain substances combine in more than one proportion,
but these proportions are rational or integral. This would be ex-
plained if there are parts between which rational ratios hold.
But such ratios imply finite quantities — though these be very
small.

(c) Certain substances combine with other substances in mini-
mum ratios. This would be explained by the hypothesis that
there are ultimate units which are the smallest that can enter into
combination with the units of other elements.

(d) Substances which can be gasified exert pressure on the walls
of the containing vessel, — the same pressure on each wall. For
equal volumes this pressure varies directly with the temperature.
These facts are explained by the hypothesis (1) that there are parts,
acting as wholes, in motion, and striking the walls of the vessel
with a certain momentum, and (2) that the velocity of each part,
and so the momentum, varies with the temperature. These parts
might be either ultimate units or simple multiples of these.

(e) Equal volumes of gases at the same temperature exert differ-

1 Cf., e.g., Jones, H. C., Elements of Physical Chemistry, 1907.

EVIDENCE FOR ATOMS 227

ent pressures. This would be explained if the gases consist of
moving parts, — the same in number in all gases of equal volumes
at the same temperature, — which are of the same mass in any one
gas, but of different mass in different gases. Then the velocity
would be the same. These different masses might be in the same
ratio as are the combining proportions of the substances in ques-
tion.

(/) Equal volumes of gases at the same pressure are of different
temperatures. This would be explained if the gas consists of
moving parts, the same in number in all gases of equal volume at
the same pressure, but with motions of different velocity. The
masses would be the same.

(flf) The data of (e) and (/) are together explainable only if in
different gases at the same temperature and pressure both the
masses and the velocities of the parts are different. The velocities
must be if the masses are. And the masses are different — as
' combining proportions ' show.

(h) That the volumes, pressures, and temperatures of two gases
should be the same, but the masses and velocities different, is
explainable, if equal volumes of gases, at the same temperature and
pressure, contain the same number of parts, either ultimate parts,
or complexes acting as units. (Avagadro's hypothesis.)

(i) That equal volumes of gases (in certain cases), at the same
pressure and temperature, combine in definite proportions is ex-
plainable by the same hypothesis.

These examples of the actual analyses which are made by way
of establishing the atomic and molecular theories are sufficient to
make my point clear. The data cited are themselves analytical
results. They involve the distinction between volume, pressure,
temperature, between different substances,1 etc., etc. But these
specific distinctions are seldom attacked. The analyses cited and
their results lead to or demand another analysis, that of a pure
chemical substance into parts called atoms, or, all taken together,
they are this analysis.

1 Cf. Section IV of this essay.

228 DEFENSE OF ANALYSIS

Put in the form of our schematic syllogism, the data discovered
by analysis, that is, the data of definite proportions, of pressure,
etc., are a set of ' existential q's.' They are facts, and in that sense
are affirmed or asserted. But their assertion carries with it the
assertion of the one hypothesis which, in the present state of our
knowledge, is the only one that will explain them all. Any other
hypothesis is excluded, since no other one hypothesis will explain
the manifold of data, found independently, and to be explained.
Then the entities, the atoms, which that hypothesis denotes, are
to be accepted as real in exactly the same sense as are the data
which they explain.

But this outline of that analysis which leads to the atomic
theory, and, as the realist would say, to the discovery of atoms, is
typical, as is also the kind of whole which is analyzed. This kind
of whole is found in any actually existing pure chemical substance.
Such a substance or whole is made up of particles, of molecules,
of atoms, and perhaps, finally, of electrons. A current of elec-
tricity, the cathode rays, the «, /3, and y emanations from radium,
etc., would also seem to be such wholes — wholes whose parts are all
alike. And the analysis outlined is typical of the various specific
analyses which are made. The existence of various data found
by experiment is explained only by the existence of certain parts.

Now in the case of certain analyses previously considered the
typical argument has been advanced against their realistic inter-
pretation, that the parts or terms to which analysis leads are the
contradictories of the whole, and that accordingly the analysis
falsifies. This attack has been found to fail in every case so far
examined. Does it fail here also in the case of atoms, electrons,
etc.?

The reply is, that in this case there is scarcely opportunity to
make this typical attack. Molecule, atom, electron do not seem
prima facie to be the contradictories of the whole of which they are
the parts. They may be made so, artificially, of course, for any
two terms, A and B, distinct from one another, can be thrown into
the contradiction-mold, A and non-A. But this process generates

ATTACK AND REPLY 229

no difficulties, for it presupposes the very absence of difficulties in
the data with which it starts. A and B are compatible, though
different, and, therefore, so are A and non-A, if non-A = B. Con-
tradiction must have a fulcrum in order to be damaging, it must
turn on some eccentric.

But, although not contradictory to the whole, molecule, atom,
electron, may be different from the whole of which they are parts.
In fact there is evidence that they are, especially in some cases.
But this is not damaging either. It shows neither that the analysis
is false, nor that it is inadequate — although it may be incomplete,
or may in fact be in error, — to be corrected subsequently. The
latter possibility must, of course, be admitted of any analysis.
Consider the actual state of affairs which is revealed by this modern
physico-chemical-electrical analysis. There are the wholes to be
analyzed — pure chemical substances, let us say. And there are
the parts revealed, — particles, molecules, atoms, and electrons.
But there are also the organizing relations. These are revealed
by the analysis just as much as are the terms. These relations
may at the present time offer many further problems, they may
be of many different kinds, they may not be the ultimate relations
involved, but at the present stage of scientific development they
are the relations which are revealed, whose field is the terms dis-
closed by the analysis, and which, together with these terms, con-
stitute the whole. The terms in relation may not constitute the
whole in an additive manner ; few wholes are so constituted ; few
properties so result. They may not be the whole, for example,
in quite the same way that points related asymmetrically and
transitively are a line. But they are the whole in some of its dis-
tinguishable aspects, just as the points in relation are the line, and
they are not the whole in other aspects. The whole has properties
which are different from the parts and the relations. Analysis,
then, by subtraction if you will, reveals these properties. The
specific properties of the whole plus those which are the terms in
relation, plus, of course, the relations, exhaust the whole. The
analysis, then, is adequate. But it may not be complete. Pos-

230 DEFENSE OF ANALYSIS

sibly it can be driven further. But, on the basis of what it is now,
it is adequate, and certainly presents no opening for the introduc-
tion of the claim that it leads to contradictions and so falsifies.

At the present stage, then, in the development of science, those
entities, such as electrons, atoms, molecules, etc., and the relations
between them, which together exclusively account for certain ex-
istential phenomena, must be accepted as existing in quite the
same sense as do the entities which they explain.

IV

PERCEPTUAL AND CONCEPTUAL ANALYSIS

WE now come to the third kind of whole, that whole, namely,
which, being itself a class, is analyzed into subordinate classes.
Is this analysis falsification ?

Let us take an example. Fluorine, chlorine, bromine, and
iodine are grouped together as the halogen elements. They are
so called, of course, because of the similarity of their compounds
with those of chlorine, such as NaCl, common salt.1 Halogen,
is, then, the class-concept. The actually existing things are spe-
cific quantities of fluorine, chlorine, bromine, and iodine — actu-
ally existing in some specific place in each case. The class-con-
cept, halogen, is none of these. But neither are fluorine and
chlorine, as class-concepts, themselves fluorine and chlorine.

Let us consider first the actually existent entities, the real
fluorine, chlorine, etc. Specific quantities of chlorine would, as
specific, be distinguished from one another, — analyzed, if you
will, from one another, and yet they would be found to have com-
mon characteristics by virtue of which they are all chlorine, and
quite similarly with specific quantities of fluorine. Yet there
is another distinction which the actual analysis makes, for ex-
ample, that between a specific quantity of chlorine and one of

1 Modern chemical investigation shows that these substances have further
properties by virtue of which they belong together in Group VII of ' The Periodic
Law.' Strictly speaking, they are sub-group A of this Group.

PERCEPTION AND CONCEPTION 231

fluorine. Put two such quantities side by side in the laboratory
under certain conditions, and make suitable tests and observations.
The two quantities are spatially distinct. But they are also found
to be qualitatively distinct in many ways, at the same time that
they are similar. Now it is this type of analysis that I wish to
examine, the type by virtue of which different kinds of entities
are discovered. It is this analysis, of course, by virtue of which,
in both common sense and science, the great manifold of kinds of
things, events, etc., are classified and systematized. Is this analy-
sis attacked ? We have only to turn to M. Bergson's ' Creative
Evolution ' to find an affirmative answer : " Our perception, whose
role is to hold up a light to our actions, works a dividing up of
matter that is always too sharply defined, always subordinated
to practical needs, consequently always requiring revision. Our
science, which aspires to the mathematical form, over-accentuates
the spatiality of matter." 1 "What is real is the continual change
of form : form is only a snapshot view of transition." "Our per-
ception manages to solidify into discontinuous images the fluid
continuity of the real." 2 Manifestly this attack is made from
the standpoint of a position that is obtained by considering the
predominance of dynamic concepts in modern physical science.
Most 'things' are changing, either rapidly or slowly. That is to
be admitted. M. Bergson draws his evidence from many such
sources. And finally, by so doing, he arrives at the position that
everything without exception is change, flux, evolution, with such
an interpenetration of parts (if there are parts) that there are no
lines of separation, that there is only One great viscous or mobile
fluid.3 With this the case, it is clear that his attack on perceptual
analysis comes under the usual form. The whole is a continuous,
flowing, trembling jelly. Perception introduces discontinuity, rest.
It selects this object, that object, this quantity and that, and
makes them static. Then it falsifies, serve though it may the
practical purposes of our action. This is the attack. Can it be
met?

1 Ibid., 206. 2 Ibid., 302. » This essay, Section I.

232 DEFENSE OF ANALYSIS

I think that it can be, and, indeed, hi two ways. First the
major position from whose standpoint the attack is made is con-
structed by applying the constitutive theory of relations to any
and all relations between any and all things.1 If there is a uni-
versal interpenetration of all things by virtue of their being re-
lated, then there are no things, no thing is any one thing, nothing
is itself, and everything is something else. But the fallacies and
difficulties of this view we have previously examined. Change
may be very prevalent, although not universal. Or it might even
be admitted for the sake of the argument to be universal for exist-
ing entities. But from that admission it does not follow that
there is only One Change or Evolution with no place at all for
typical and individual changes. Conceivably there might be only
One Change, qualitative and quantitative, or there might be many
changes, qualitatively and quantitatively distinct. Then analysis,
discovering which is the case, would be valid; for the analysis
which I am discussing here is not limited to 'getting at' the
statical. It can also get at the dynamical, the evolving, the chang-
ing quite as well, as has been shown. Bergson, for example, him-
self tacitly admits and presupposes this in distinguishing three
kinds of changes which we perceive and for which we have con-
cepts, namely, qualitative, evolutionary, and extensive.2 Every
opponent of analysis or of pluralism admits it in so far as he uses
concepts denoting different kinds of change.

Accordingly it can be admitted that the things, the qualities,
etc., which we perceive and distinguish in concrete cases are not
genuinely statical, but are changing, slow though this process be,
and yet that this perceptual analysis is quite valid. Different con-
crete entities, whether or not further analysis show them to be
processes or just plain ' things ' — equilibriums if you will — can be
perceived, and they can be perceived as different, whether the dif-
ference be only a spatial and temporal one, or also a qualitative
one. In the former case the two or more entities are perceived
as instances of the same qualitative complex, and the possible

1 Se« Section I. * Op. rit., 303.

CONCEPTS 233

existence of still other cases is implied. In the second case it is
implied that other instances, similar respectively to each of the
two or more qualitatively different complexes, may be found.
Then there are two sets of cases, each case individual, but each
set also differing as a set from the other set. The individuals of
each set are simply different spatially and temporally, one or both,
but they are individuals or instances of a certain complex of quali-
ties, other instances of which can be conceived. But in any case,
whether found to exist or only conceived, they differ qualitatively
from the individuals of another set.

All this is trite. It is but another way of saying that among
certain individuals which are numerically distinct, qualities or
groups of qualities are found which are numerically the same, and
of which other instances can be conceived. There are qualita-
tively different ' states of affairs ' of different groups of individuals,
the individuals of each group forming that group by virtue of
being instances of the same state of affairs. Then there are states
of affairs which differ qualitatively and are themselves numeri-
cally distinct. Briefly, there are different concepts revealed by
analysis. What, now, is the character of these states of affairs ?
Briefly, it may be said (1) that the state of affairs, the concept, is
not the printed or spoken sign, the word. It would subsist, did
the signs not exist. (2) It is not the knowledge or idea of the
state of affairs, for again there would be a real state of affairs even
if it were not known. (3) It is not identical with the individual
cases, whatever these be. Number is not any one number, man
is not a man, etc. (4) It is not necessarily even physical or mental,
even when the individual cases are physical or mental existents.
Thus the state of affairs, indivisibility, is not itself an indivisible,
nor is mentality itself mental. Further, there are states of affairs
of entities which are neither physical nor mental, that is, which do
not exist; for example, arithmetical continuity.

Analysis, so-called conceptual analysis, reveals, therefore, quali-
tatively different and numerically distinct states of affairs or con-
cepts. They may be the states of affairs regarding change, for ex-

234 DEFENSE OF ANALYSIS

ample, the laws of change, as well as regarding entities which, like
points, do not change. Is, now, the analysis which gives or is iden-
tical with these concepts open to attack ? Suffice it to say that it
is attacked. Says M. Bergson, "Concepts are outside each other,
like objects in space; and they have the same stability as such
objects, on which they have been modeled." "They are not the
perception of things, but the representation of the act by which
the intellect is fixed on them." "They are, therefore, symbols." l
"The intellect is not made to think evolution, in the proper sense
of the word — that is to say, the continuity of a change that is
pure mobility." 2

The attack here is in part at least a variation of the typical at-
tack which we have found to be made in every case so far consid-
ered. Concepts are characterized as statical entities, solid-like
'things,' external to one another, etc. How, then, can they
relate or refer to that which is not statical, but is concrete, actual
process, change, evolution? This is the variation of the typical
attack. What is its central principle? Briefly, it is this, That
only like entities can be related. Therefore the concept, which
is static, cannot be related to or refer to that which is not static,
but is process, change, etc. The concept by its very nature is
inadequate to draw out or present the character of that which
is its contradictory in nature. Or, if the concept be admitted to
be the state of affairs, the law, or what-not, of a process, then it
falsifies ; it in some sense makes static that which is not.

This, then, is the attack. Can it be met ? It undoubtedly can
be, and first by an argument both reductio ad absurdum and ad
hominem. Any party making this specific attack invalidates his
own attack and tacitly accepts the validity of conceptual analysis
in talking about evolution, process, and change. These are con-
cepts, and whether they are statical or dynamical, they are cap-
able of meaning and of referring to that which is — well, just
what they designate, namely, change, evolution, etc. If the con-
cept be dynamical, then we have like entities related ; if it be static,

I0p. tit., 160-161. * Ibid., 163.

ATTACK AND REPLY 235

then unlike things. But in either case it is related to that of which
it is the concept. The attacking party tacitly grants, then, that
there is nothing in the nature of a concept which prevents it from
being a concept or state of affairs of other entities, whether these
be like the concept or not.

With this point cleared, it can also be readily seen that on the
basis of a similar argument the attacking party can be forced to
grant the validity of that further conceptual analysis which is
identical with distinguishing different states of affairs. He him-
self distinguishes different kinds of change. Then he admits that
the general principle of having kinds is valid. The further analy-
sis of some of these kinds may be difficult. But the analysis of
the larger whole into these types is quite valid. In fact, it is
presupposed by the attacking party to be. Then further similar
analysis may also be. I conclude, then, that this specific attack
is not to be taken seriously, but that it is made under the influence
of analogy and of misleading figures of speech.

The other aspect of the attack on conceptual analysis is not a
variation from the type. It is practically the same as the attack
on perception. Perception, so it is held, falsifies by breaking up
into statical and sharply separated parts that which is really one
concrete universal change. Conception, it is also held, falsifies
by making statical that which is not. Here the defense, then,
is the same as against the attack on perception. The validity of
conceptual reference and analysis is presupposed by the attacking
party himself in his distinction of different kinds of evolution and
change. Each of these kinds has many instances. Each would
be a genuine state of affairs, which, though related to other kinds,
would be just that kind. Then he has accepted the principle that
valid conceptual distinctions or analyses can be made. That is
all that is necessary to refute his attack, for his attack is on the
principle, and not on the details of any specific conceptual analysis.

I conclude, therefore, that conceptual analysis qua concep-
tual is quite valid. Just as there are individuals, whether these
be things or processes, which can be validly distinguished in per-

236 DEFENSE OF ANALYSIS

ception, so there are types of individuals, which types, although
they are similar to one another to a greater or less degree, are also
different and distinct. The individuals may be of the most vari-
ous kinds ; they may be physical or mental existents, complexes or
simples, relations or terms, motions, or rests, or qualities, or things;
they may be entities which, like the numbers, are neither physical
nor mental, but are subsistents. Thus we are already distinguish-
ing types. But, whatever be the concrete individuals, the state of
affairs, the type, is, as a state of affairs, only subsistent, and is iden-
tical with itself. It is just that state of affairs, whether individ-
uals corresponding \o it exist now or not; it retains its meaning as
these individuals may come and go. Further, as a specific state of
affairs, it may differ from other states of affairs ; in fact, although
related to them, it may be quite independent of them. What the
relation of different states of affairs is, is for analysis to determine.
That is the problem which the attacking party always admits to be
solvable, although his solution is different from that of the defend-
ing party. For the latter, analysis solves this problem, with the
outcome that certain types are quite independent of others. Thus,
numbers as a type are independent of instants, of points, of material
particles, of acts of counting, in the sense that all these imply the
numbers, but not conversely. There is a certain hierarchy of types.
Certain types could subsist, or exist, without others, though not
conversely.

In summary, then, I conclude that that kind of analysis in
which the whole analyzed is a type or class with subordinate
classes, or finally a class with individuals as terms, is not invali-
dated by the attack, but remains a method whereby entities are
discovered which are as real, and real hi the same sense, as are those
wholes to which the attacking party alone attributes reality. "

237
V

THE ANALYSIS OF ORGANIC WHOLES1

I NOW come to the examination of the analysis of the fourth
kind of whole, illustrated by any specific finite quantity of any
specific chemical compound and by an organism. The character
of the analysis which I wish to defend, as well as the character
of the wholes analyzed, will be made clear by a definite example,
that of the analysis of water. Water is shown by analysis to be
a compound. As water, it has certain properties, some of which
are found in other compounds, others not ; in the latter case the
properties are specific. All the properties are classified as either
chemical or physical. The chemical properties are those which
are involved in the fact that water reacts or combines with certain
other compounds and with certain elements. Among the im-
portant physical properties are specific gravity, refractive power,
boiling point, electrical conductivity, absorptive power, and
elasticity. Water as a whole has these two types of properties.

By electrolysis, however, and by some other supplementary
modes of experimentation, water is actually split up or analyzed into
two substances, hydrogen and oxygen, which are gases under normal
conditions. Investigation of these shows that each has many of
the same kind of properties as has water, although with numerical
values different from the values in the case of water. Properties
common to hydrogen, oxygen, and water are all the above enu-
merated physical properties. But the chemical properties of the
three substances are different. However, in the case also of the
physical properties, the numerical value of certain properties of
the whole, water, namely, the last four, is not simply the additive
result of the values of these same properties in the parts. In the
case, then, of both kinds of properties, chemical and physical, there
would seem to be something in the whole which is not in the parts,
and conversely. If the whole be experimentally synthesized out
of the parts, then something new appears as properties of the

^ Cf. Pitkin, this volume.

238 DEFENSE OF ANALYSIS

whole, something which is new qualitatively as well as quanti-
tatively. On the other hand, if an experimental analysis be made of
the whole, then the whole is also found to have properties which
the parts do not have. These properties are put ' in relief ' by the
analysis; they are a residuum, characteristic of the whole as a
whole, and revealed by an analysis which at the same tune reveals
the parts or elements, and, through its ramifications, the organiz-
ing relations.

This fact, that hi the actual synthesis, artificial, or natural
and developmental, of existential wholes out of parts, new prop-
erties or new values appear, is a matter of great importance. It
is a fact, too, which is accepted by authoritative investigators.
Says Professor Nernst,1 "A large number of physical properties
have been shown to be clearly additive; that is, the value of the
property in question can be calculated as though the compound were
such a mixture of its elements that they experience no change in
their properties." Examples are the volume, refraction, magnetism,
and heat of combustion of organic compounds. But other prop-
erties are not additive. "The kind of influence of the atom in a
compound is primarily dependent upon the mode of its union, that
is, upon the constitution and configuration of the compound. Such
non-additive properties are called constitutive." Examples are
the absorption of light, the rotatory power, the melting point.

Now modern physics and chemistry, physics chiefly, carries
this typical analysis of water further; it analyzes the two con-
stituents of water, hydrogen and oxygen, to discover that these
in turn consist of parts related in certain specific ways. In fact,
at no stage of this physico-chemical analysis are the organizing
relations unrevealed. Hydrogen and oxygen, and presumably
all the elements, are shown to be composed of electrons, that is,
of negative electrical charges hi a positive electrical field. Further
analysis shows that these electrons are in very rapid orbital mo-
tion, and that their mass is a function of their velocity. The
atom turns out, then, to be a mechanism, — an electro-mechanism,

i Nernat, W. (Lehfeldt, R. A., trans.), Theoretical Chemistry, 365.

NEW PROPERTIES 239

however, just as the molecule is a mechanism whose parts are
atoms, and the particle a mechanism whose parts are molecules.
But with the atom an electro-mechanism, the analysis can be
pushed further. The electron is itself an electrical field of force.
It is, then, a three-dimensional manifold of elements which are
intensity-points forming an ordered series. But the electrons
are in motion. Then their motion can be analyzed after the
manner previously presented.1 Clearly, various laws, arithmetical,
dynamical, mechanical, as they have previously been presented,
have an application to the entities within the molecule. From
these laws, however, no specific molecule with its specific char-
acteristics can be deduced, yet, conversely, any specific molecule
is found to imply these laws.

However, we do not need to go so far as to take into considera-
tion the analysis involved in these laws in order to establish an
important point. Let us stop with the analysis of the atom into
electrons, and compare the properties of the latter with those
of the former. Electrons as individuals have certain properties
which atoms have, namely, a volume, a specific gravity, a mass,
an attractive power (whatever this may be), but they lack refrac-
tive power, rotatory power, electrical conductivity, absorptive
power, which properties the atom has. Similarly, any finite quantity
of atoms, as atoms and not as molecules, for example, vaporized mer-
cury, has, as a whole, properties which the individual atoms do not
have, and likewise with the molecule. At each stage in the syn-
thesis of wholes out of parts which are in turn wholes until we get
to the intensity-points of that field of force which is the electron,
there are properties of the whole which are not found among the
properties of the parts. But analysis reveals what these wholes
are, what their parts are, what the properties of each are, and
what the organizing relations at each level are. It allows for a
whole which is not merely the sum of its parts, and which, with its
properties, cannot at the present stage of science be deduced from
those parts. Yet it also allows for the empirical ascertainment in

1 This easay, Section III.

240 DEFENSE OF ANALYSIS

many cases of the functional relation between the properties of
the whole and those of the parts.

Can this analysis be successfully attacked? The question
is most important, for it is just this kind of analysis that is
identical with the experimental procedure of many sciences. Thus
would physiological chemistry analyze organisms and their parts ;
thus do chemistry and physics proceed.

The attack is undoubtedly made. However, there is no oppor-
tunity for introducing that typical attack which we have consid-
ered elsewhere. Here, part and whole do not seem to be contra-
dictories. The only opportunity for attack, then, consists in the
claim that, in a whole which is constituted by interpenetrating and
causally interacting parts, no part can be experimentally re-
moved without altering it.1

The reply is that perhaps it cannot be — in the instance of the
kind of whole we are now considering. There do seem to be wholes
which consist of parts, which, related, do modify or influence one
another, or which, perhaps, are constituted by virtue of their
relation to one another. At least, this is one hypothesis. Such
wholes are well illustrated by an organism, and for this reason are
called organic. As wholes, they do seem to have properties which
the parts have not, properties which are not derived additively
from those of the parts. Let the parts be brought together in a
natural process, and we have both creation and evolution. Let
them be brought together in the laboratory, and we have creation
accompanying synthesis. In either case there is a creative synthe-
sis, natural or artificial. But organisms are not the only kind of
whole which presents this synthesis. Non-living things do also,
quite as well. Chemical compounds do, atoms do, every physical
complex does. Both part and whole in every case may be arithmet-
ical, dynamical, mechanical, etc., but in each case also they are
more than this in that, for example, the mechanical laws which
apply are limited by the peculiar qualitative constants which
render each whole and part specific.

1 This essay, Section I.

TWO HYPOTHESES 241

With reference, now, to the actual experimental analysis of
these organic wholes there are two hypotheses, each of which is
compatible with the realistic interpretation of analysis. The one
hypothesis is, that when the parts are analyzed out they are changed
in certain respects. This would not necessarily be the case with
all the properties of the parts. For example, in the case of water,
certain properties, such as the specific gravity and refractive power
of hydrogen and oxygen, would seem to be the same, both when
these substances are in and out of combination. But with other
properties, especially the chemical, the case is different. Some of
these seem to be gained by the parts when analyzed out, and to be
lost when the substances are combined, with the supplementary
gain by the whole. Analysis and synthesis are, then, comple-
mentary processes. Each also is a natural process. Both create.
In each, relative to the other, new properties appear, and there is
a genuine creation which is not explained away by saying that
that which appears as new has been potential all the time. By
this first hypothesis, then, the parts when in situ in the whole will
be actually constituted by virtue of their relation to other parts.

By the other hypothesis, the parts, whatever these may be,
electrons, atoms, molecules, particles, remain quite the same,
quite unmodified whether they are in or out of the whole. But
then, to accord with fact, it must be granted that new properties
arise for each successive whole. By this hypothesis, with elec-
trons combined to form an atom, the electrons remain just what
they were before they were combined, yet the combination as
a whole has properties which the electrons lack, and similarly
with the combination of atoms into molecules, molecules into
particles, and so on upward, as it were. Both hypotheses, however,
recognize a non-rational element in nature, — at least so far as our
present knowledge goes. The time may come when the new
properties of the whole can be deduced from those of the parts,
but at the present time this deduction is impossible, and it is an
open question as to whether this impossibility is due to the struc-
ture of existence, or to our ignorance,
a

242 DEFENSE OF ANALYSIS

Both hypotheses are, however, perfectly consistent with the
realistic interpretation of analysis. In neither case does analysis
lead to parts which are the contradictory of the whole. The analy-
sis is itself a process, an event. By the first hypothesis, that which
is taken out, or put in, is altered. But in each case analysis reveals
what the alteration is, and in each case the resulting properties
are real. Analysis is itself simply a process from which real prop-
erties result, by which real properties are changed.1 By the second
hypothesis, that which is taken out and put in is not altered, but
remains the same entity. Yet when this synthesis takes place,
the parts become related as they were not before, new organizing
relations are instituted, and a whole with new properties is formed.
But here also analysis reveals these properties, the organizing re-
lations of the parts, and the parts with their properties. Both
analysis and its complement, synthesis, are processes, the former
revealing parts which remain what they are even when their re-
lations change, the latter instrumental in causing new properties
to arise. Neither, however, furnishes any opportunity for the
claim that either the parts revealed, or the properties and whole
produced, are not to be interpreted quite realistically, or as in any
sense contradictory and so falsifying. Change is a fact. That
it is, is one of the attacking party's chief arguments against analy-
sis. Then most assuredly that party cannot consistently attack
experimental analysis on the ground that it itself is a process and
brings about changes. The attack could be made only as regards
the character of the changes instituted, and this is, as we have seen,
the principle of the typical attack on analysis — that it changes
the whole to parts which are the whole's contradictories. But
here there is no opportunity for this claim. Molecules are not
the contradictory of particles, nor are atoms of molecules, nor
electrons of atoms.

I conclude, then, that, like the other analyses examined, the an-
alysis of organic wholes stands unimpeached by any attack which

1 It has been previously shown, of course, that change presents no insuperable
difficulties to analysis. Section III.

ANALYSIS OF ORGANISMS 243

has as yet been made upon it. These wholes are, more clearly
than any others which we have considered, not simply the addi-
tive result of their parts. They have new properties, properties
which the parts lack. All physical and chemical wholes, both
living and non-living, are of this sort — excepting, of course, the
ultimate simple. The analysis is adequate. It reveals the parts,
the organizing relations, and the properties of the wholes them-
selves. Valid also is this analysis. It may change that which is
analyzed out, as by the first hypothesis, or it may not, as by the
second. But in either case the part is to be accepted at its face
value in accordance with the fundamental realistic postulate.
For there is nothing in change qua change which runs counter to
this postulate.1 Change qua change does not prejudice the real-
ity of either the terminus a quo or the terminus ad quern. That is
admitted both tacitly and explicitly by the opponents of analysis.
Then they cannot hold a brief against analysis on the ground
alone that it involves change — as this specific kind of analysis
does in distinction from analysis in situ.

ORGANISMS AND THEIR ANALYSIS

Organisms are the kind of whole, and their analysis is of the
type just presented. The recognition of this throws much light
on the question as to what the nature of the organism is, what the
nature of life is. It clarifies the issue between the so-called vitalist
and mechanist in biology. All the evidence, now, shows that the
organism consists of cells, of colloidal particles in solution, of
molecules, of atoms, and of electrons. At each level, as we go
upward synthetically, new properties appear. Going downward,
analytically, there is on the whole a loss of properties. There is a
tendency toward simplification. In general all this is admitted
by all parties. By all, the organism is conceded to be this kind of
whole. The question at issue, however, is, is it more ? Now a
physico-chemical complex is a whole in which there are parts,

1 Cf., e.g., Bergson, Creative Evolution, and Time and Free Will.

244 DEFENSE OF ANALYSIS

molecules, atoms, etc., each one of which is in turn a whole. Well-
known laws apply to these various wholes and parts, but only as
qualified or limited by certain ' constants ' expressing the numerical
value of the properties at each level. Thus, for example, the
principle of the Conservation of Energy, the principle of D'Alem-
bert, the Second Law of Thermodynamics, the Laws of Motion,
etc., apply to the physico-chemical complex, but only under con-
ditions which are expressed in the formulae in such a way as to
give, by integration, values which are confirmed experimentally.
In this sense, then, the complex is a mechanism. But it is not a
pure mechanism, that is, its laws are not those purely mechan-
ical laws which are obtained by eliminating all constants by
successive differentiation. But it is mechanical in the sense
that the electrical current is mechanical, namely, with purely
mechanical laws qualified by limitations obtained by measuring
electrical phenomena. All this can be granted, and the ad-
mission still be made, that every chemical compound is peculiar
and specific, different in some respects from every other compound,
and with properties, as a whole, which do not characterize its
parts.

Is the organism, the individual organism of any species or vari-
ety, plant or animal, anything more than just such a specific
physico-chemical complex, specifically different, of course, from
other physico-chemical complexes which are organisms?

Vitalistic theories are not many. One traditional vague theory,
that of the older vitalists, holds to the existence in an organism of
a vital force or energy, but this entity has never been discovered
experimentally. However, quite evidently, did it exist, it would
not make the organism non-mechanistic. For energy is subject to
mechanical principles. A vital energy would at best add only one
more mechanistic element to that complex which is the organism.
Another theory makes a mental factor universally parallel with
the physiological factors. Then, if the latter be mechanistic
(which is the question at issue), the former, as parallel, is also. A
third theory, distinctively non-parallelistic, places a psychical

VITALISTIC THEORIES 245

entity 'in control/ as it were, of the discharge of that potential
energy which, according to the theory, it is one of the distinctive
features of the organism to store up in greater amounts, perhaps,
than do inorganic complexes. Thus various ends or purposes,
'entertained' by this entity, could be accomplished by vaiying
the direction of the release of the potential energy. Once released,
however, the specific energy discharged would take place in ac-
cordance with the usual principles of inorganic events. Now this
is a theory which, if it be true, does mean a real difference between
the organic and the inorganic, for by it the organism as a whole
would present in its behavior a range of variations under the same
conditions which it would not do if it were simply and only a
physico-chemical complex. Under the same conditions the or-
ganism would now do one thing, now another. The holding of
this theory to be true would have an important influence on the
biologist's attitude. It is, in fact, nugatory of scientific biology,
and if the theory were true, vitalism would have a distinctive
'meaning, differentiating it from mechanism. However, the
theory is not supported by facts. The organism, even in the
case of those phenomena which, like regenerations, restitutions,
variable morphogenesis, etc., are held by some to support a
vitalistic theory, is found to do the same thing under the
same circumstances. That, however, which these phenomena are
held by others to show is, that under different circumstances the
same end or outcome is accomplished or gained. This gives a
fourth theory, according to which there is hi every organism a
psychical entity which, in the midst of varying circumstances, suc-
ceeds in bringing about a definite end.1 This is the hypothesis.
However, the physically observable fact is, that that which by
the hypothesis is an end is prima facie a later stage in the develop-
ment or behavior of the organism. It may also be more than this,
that is, even as it would be interpreted by the hypothesis under

1 This is Driesch's Neo-vitalism and doctrine of the Entelechy. Driesch, Hans,
The Science and Philosophy of the Organism, Gifford Lectures, 1907, 1908, and
other works.

246 DEFENSE OF ANALYSIS

consideration, it may be a consciously held and purposed end,
aimed at now, and accomplished in the future. But even if this
were the case, its primary status as ' later stage ' would not be
altered thereby. In fact, to be a ' held ' end that is accomplished,
it must also be at least an end in this sense, namely, a 'later
stage ' ; it could be the latter without being the former, but not
conversely. But if the end be an end in this sense of ' later stage/
then, whatever else it may also be in certain cases, it is the kind
of end that inorganic physico-chemical complexes also present.
In general, among these complexes, the same effect, or what
within certain limits is the same effect, can be produced by many
causes. Thus a specific chemical compound can be synthesized in
many ways. Further, as wholes, these inorganic complexes present
properties which the parts do not have; they are mechanisms,
though not machines. These two features are, now, so far as the
ascertainable physical facts are concerned, all that the organism is.
If, however, a special entity, like Driesch's Entelechy, were brought
in to explain these facts in the case of inorganic complexes, not
only would it be superfluous and not explain, but the facts them-
selves would not be altered thereby. The complex would be quite
as mechanistic with the entity as without it. But for these com-
plexes there is no necessity for bringing in such a special entity.
The only difference which its presence would make would be the
difference between presence and absence.

If, then, vitalism is defined in accordance with the fourth hy-
pothesis, it is a position which is meaningless in the sense that it
does not succeed in making that distinction between the organic
and the inorganic which it aims to. For if such a special entity be
put into living things, there are the same reasons for putting it into
non-living. But with either its presence or its absence common to
both realms, no distinction between these can be drawn or found
on its basis. But this is not to say that a consciousness or aware-
ness of some specific character may not be admitted to arise in
certain organisms under definite conditions. It may. But such
an awareness does not demand a vitalistic interpretation of the

VITALISTIC THEORIES 247

organism. It does not explain the end in the sense of ' later stage,'
but only makes of this a ' held ' or ' purposed ' end, to be accom-
plished in the future. The ' later stage,' as well as the ' creative
synthesis,' as actually occurring and so observable, is explained
and accounted for sufficiently in connection with and in relation
to earlier stages. The hypothesis of a supervening awareness adds
no explanatory element to this account. This awareness may
occur, and it is good realism to admit that it may, but, if it does,
it is to be distinguished from a special entity like an ' entelechy,'
which is held to persist and control and direct, and to explain
both accomplished end and 'creative synthesis.'

A difference in specificity is, then, the only difference between
the living and the non-living, but this difference furnishes no ground
for holding a vitalistic as opposed to a mechanistic theory.1 Those
vitalistic theories which by definition do mean something different
from mechanism are not confirmed. Others only add an hypo-
thetical entity which makes no difference to the facts discovered.
Both realms, the organic and the inorganic, are mechanistic at the
same time that they are specific. This means that neither realm
is purely mechanical. Each furnishes an opportunity for applying
mechanics by introducing into mechanistic formulae the constants
found by measuring the specific properties of each complex.
Therefore, if this view be called vitalism, there is no difference
between vitalism and mechanism. Vitalism is but another
word for not-pure mechanism. Both theories mean, of course,
that the organism can be analyzed experimentally and formally.
I cannot agree then with M. Bergson's attack on the analy-
sis of the organism or with that peculiar vitalistic theory of his
which makes all complexes unanalysable, and which would make
everything vitalistic.2

1 Cf. a discussion between Messrs. Hitter, Jennings, and Lovejoy in Science, 1911,
34, Nos. 847, 851, 857, 859, 864. For a fuller exposition of the position taken in
this section of this essay, see my paper, The Energy of Segmentation, /. of Exp.
Zoology, 4, 2, 284-315.

* Creative Evolution, 162 and 225 et passim.

A REALISTIC THEORY OF TRUTH AND ERROR

A REALISTIC THEORY OF TRUTH AND ERROR
BY WM. PEPPERRELL MONTAGUE

INTRODUCTION

REFLECTION upon the fact of error has been the principal
cause for the abandonment by philosophers of the standpoint of
natural or naive realism. According to that view, consciousness
is conceived after the analogy of a beam of light which reveals the
nature of the very world of which it is itself a part. And as the
objects revealed by a light in no sense depend upon it but rather
does the shining of the light depend upon them, so do the objects
revealed in consciousness in no sense depend upon consciousness
but rather does the occurrence of consciousness depend upon them.
When it is found, however, that some of the objects revealed in
consciousness — such as the events of a dream — have no place of
their own in the spatio-temporal system of interacting beings,
but appear to be active only in the individual who experiences
them, there arises a doubt as to whether consciousness ever directly
reveals any other obj ects than the states of the one who is con-
scious. Instead of being viewed as analogous to light, conscious-
ness is now regarded as analogous to a photographic plate on which
objects external to the knower are represented or symbolized by
the 'ideas' which they produce. This epistemological dualism,
however, becomes unsatisfactory as soon as it is realized that we
can ascribe to the external objects inferred as the causes of our per-
cepts no locus or nature other than that of the percepts themselves.
Because of this, the copy theory of knowledge gives place to the
theory of epistemological idealism or subjectivism, according to
which the world in which we live is conceived as a product, fash-
ioned by consciousness from the raw materials of its own states.
The internal contradictions of each variety of this third theory,

251

252 A THEORY OF TRUTH AND ERROR

the manifold difficulties in the way of reconciling any form of it
with the procedure of common sense and of science, and, finally,
the pathetic dependence of consciousness upon the very objects
which it is supposed to create — have brought about the realistic
revolt. And as the departure from realism was due to the sub-
jectivistic interpretation of error, so the return to realism must
be based upon a realistic interpretation of error, and hence of its
correlate truth. I shall treat the subject under three main
heads : I, The Meaning of Truth and Error ; II, Causality and
Consciousness in a World of Pure Fact; III, The Genesis of
Truth and Error.

THE MEANING OF TRUTH AND ERROR

1. Definition of True and False. — I shall use the term 'truth'
to connote 'true knowledge' and the term 'error' to connote
'false knowledge'; hence the definition of truth and error will
resolve itself into the definition of true and false. I hold that
the true and the false are respectively the real and the unreal, con-
sidered as objects of a possible belief or judgment. There is, that is
to say, the same difference between what is real and what is true
as between George Washington and President George Washington.
President George Washington refers to Washington in a certain
relation to our government. George Washington denotes pre-
cisely the same individual without calling attention to the presi-
dential relation.

2. The Meaning of Real and Unreal. — Having defined the true
as the real in so far as it is an object of an actual or possible
belief, we seem called upon to go on to define the real and to
define belief. I am not sure that we should not be justified in
refusing to comply with this demand on either one of two grounds :
(1) that 'real' and 'belief,' so far as their connotation is concerned,
are ultimate and indefinable terms and that any attempted defi-
nition would be circular ; (2) that their definition would be super-

THE TRUE AND THE REAL 253

fluous in the sense that whatever definition was adopted, the
true would always be found to be formally and denotatively
identical with the real. That is to say, I might permit the reader
to adopt his own view of the meaning of real — whatever that
might be, and then without criticising it at all, challenge him
to show any difference between what he regarded as real and
what he regarded as true. I will not, however, avail myself of
this plea, further than to say that my definition of the real, if it
be rejected, should not be taken to invalidate my thesis as to its
identity with the true.

I shall use the term 'subsistent' to denominate any one of the
actual and possible objects of thought. The subsistent, as thus de-
fined, is the only class, if class it can be called, which has no nega-
tive, as is at once evident from the fact that if we formulate in
the usual way what should be its negative, viz: 'what is not a
possible object of thought,' we have — if our words mean anything
at all — merely another 'object of thought.' In short, the subsis-
tent makes an absolute summum genus. Moreover, every subsis-
tent has as an inseparable aspect of its meaning an 'is' relation
to some other subsistent ; hence every subsistent is or involves a
proposition.1 Now there is one great group of these subsistent
objects or propositions which is easily distinguishable from the
rest. It is the space-time system. It has for its elements what I
will call ' events ' — that is, groups of qualities standing in the ulti-
mate relation of occupancy to one time and one place. This use of
the term 'event' may be criticised on the ground that what we
call events usually take a certain amount of time and hence include
in their meaning the notions of duration and change. Nevertheless
I can think of no better word than events to indicate the elemental
particulars of the existing world. These ultimate particulars must

1 That every subsistent is or involves a proposition should not be interpreted as
incompatible with the self-evident truth that the terms which stand as subjects and
predicates of propositions also subsist. I mean only that as no term subsists apart
from an identity-complex or proposition, the totality of terms denotes the totality
of propositions. I shall use the word ' object ' as equivalent to term-complex.

254 A THEORY OF TRUTH AND ERROR

be the terms, themselves changeless and durationless, between which
the relations of change and duration obtain. They are the occu-
pants of single spaces at single moments, the temporal cross-
sections of single objects. The ordinary units of physical analysis
are enduring things, spatially simple but temporally complex,
like atoms or electrons, which are spoken of as 'changing/ 'act-
ing,' 'causing,' etc. So deeply rooted in thought and language
is this usage that it is difficult to avoid, and in this article I have
often relapsed into it. And yet, but little reflection is needed to
show that we have the right and indeed the duty of dividing the
world not only 'vertically' in space into substance-units, but,
also 'horizontally' in time into phase-units.

Now because of the qualitative complexity of these event-ele-
ments, each one of them constitutes in itself at least one proposi-
tion, — proposition being taken, as above, to denote a complex, in
which may always be distinguished two terms related by some
tense and number of the verb to be.1 Thus the occurrence of an
explosion at a given time, in a given place, is an event and may be
expressed in the proposition — "The matter-and-motion complex
of qualities (called 'explosion') is what occurred in the spatio-
temporal region, Si,ti." The momentary qualitative state of a
resting or moving body could in the same way be made the sub-
ject of a proposition whose predicate would be 'occurred in a
given time and place.' Besides the events themselves and their
internal identities just mentioned, there are in the spatio-temporal
system the spatial and temporal relations between specific events.
These, too, can obviously be expressed as propositions as, for ex-
ample, "The death of Socrates is a thing that happened before the
death of Kant." When the temporal aspect of an event is not of
especial interest the proposition will be a relation not between

1 The verb to be when used as a copula in a proposition indicates absence of
duality of denotation, or position, combined with presence of duality of internal
nature, or connotation. Thus when we say, " Iron is useful " we call attention to
the fact that the connotative duality of iron and useful is combined with denotative
unity. (Cf . my article on The Meaning of Identity, etc., in The J. of Phil., Psychol.,
etc., 3, 127.)

SUBSISTENCE AND EXISTENCE 255

events but between 'enduring things or beings.' (An 'enduring
thing' may be defined as a temporally continuous series of events
which are more or less the same in quality and which do not at
any one moment occupy discontinuous regions of space.) "Cae-
sar lived before Napoleon," or "Caesar resembled Napoleon,"
would be examples of such propositions. Now the propositions
so far considered all deal in one way or another with elements
of the spatio-temporal system or with groups of these elements.
They are what are usually called existential propositions. But
there is another class of identity relations to be found in the spatio-
temporal system, viz., relations between parts or aspects of differ-
ent events. Now as all the qualities (including the generic spa-
tial and temporal qualities themselves) which are exemplified at
various times and places — may be exemplified equally well at
other times and places, the relations of identity that hold be-
tween them will have a reality that is independent of any par-
ticular existence. These relations constitute our 'universal' and
so-called 'nor -existential' or 'merely valid' propositions. Ex-
amples of such propositions would be "Orange resembles yellow
more than green" f "7 + 5 = 12" ; "What is true of a class is
true of each member of a class," etc. These relations are presup-
posed or implied by the spatio-temporal system of existent ob-
jects. So we can say that the real universe consists of the space-
time system of existents, together with all that is presupposed by
that system. And as every reality can be regarded as a true iden-
tity-complex or proposition, and as each proposition has one and
only one contradictory, we may say that the remainder of the
realm of subsistent objects must consist of the false propositions
or unrealities, particular and universal, which contradict the true
propositions comprising reality.

And now as to the definition of belief — I will say merely that
it is the attitude we take toward any proposition that appears
to be true or real, and that it carries with it a tendency to act on
that proposition. But what it means for a proposition to 'ap-
pear' to be true, and how false propositions can appear true,

256 A THEORY OF TRUTH AND ERROR

we shall have to consider in the later sections of our dis-
cussion.

3. Objections to the Definitions. The Verbal Fallacy of Psycho-
physical Metonymy. — An objection to this identification of the
true and real now presents itself. True and false, it will be
said, are adjectives which apply to beliefs, that is, to the acts or
states of an individual. We do not call objects true or false ; we
call them real or unreal — beliefs alone are susceptible of truth
and falsity. To which it may be replied that true and false only
apply to beliefs in a metonymous or borrowed sense, i.e. in virtue of
the relation of the act of believing to the object believed in.
There would be no sense in calling an act of belief as such either
true or false. It is always because of what is believed that the
belief is true or false. Belief borrows its truth or falsity from
its object or content. When we speak of a belief as true we
mean that the thing believed is a fact, is real, is so, is true.
Language is full of similar instances in which words that properly
apply only to objects in a certain relation are used as if they
applied to the relation itself. When we say — "What a fine
sight ! " we do not mean that the act of seeing or process of see-
ing is fine ; and yet the word 'sight' to which the adjective 'fine'
was applied does mean, primarily, the act of seeing. We have
used it metonymously to indicate the thing seen. I call a neg-
lect of this ambiguity the verbal fallacy of psychophysical me-
tonymy. The ambiguous use of the word 'thought' offers another
example of it. Thought means primarily the act of thinking, yet
when we say, "A penny for your thoughts," we do not desire
to know about the acts or processes of thinking that are taking
place in our neighbor's organism, but about the things or objects
of which he is thinking. As it is with 'thoughts,' 'sights,' and
'beliefs,' so it is with judgments which are the expressions or
utterances of beliefs. Judgment, like belief, is originally a
name for an act or process on the part of an individual ; both
words have come also to be used to^denote what is believed and
what is judged,/ i.e. the proposition or identity-complex asserted.

VERBAL FALLACY OF METONYMY 257

Truth and falsity never attach to judgments or beliefs in the
first sense, but only in the second. In other words, they attach
to propositions or objects, and not to judgments as acts. If
the thing believed is a fact, the judgment expressing it is true ;
if not a fact, then the judgment is called false. But in case any
doubt on this point still remains in the reader's mind, let me put
this question : If we wished to know whether certain beliefs
that we held about the properties of triangles were true or false,
whom should we consult? The psychologist? Certainly not.
We should go to the mathematician. But why? The psycholo-
gist is supposed to be an expert on mental processes, and if the ad-
jectives true and false were to apply to -beliefs as mental processes,
he would be the one to settle our difficulties. We should go to the
mathematician, however, because our desire to know whether
our beliefs about triangles were true or false could be satisfied
only by one who knew about triangles. So with all cases of doubt
as to truth and falsity, we go to the person who knows about the
things believed rather than to him who knows about the processes
or acts of believing. That true and false apply to belief in the
objective sense will appear still more clearly if we realize that there
are many other adjectives that apply to belief in its subjective
sense. If, for example, instead of being interested in the truth
or falsity of a belief we were interested in whether it were com-
forting, inspiring, or healthy, then we should, very likely, go to
the psychologist rather than to one who was an authority on
the subject matter of the belief.

4. First Consequence of the Verbal Fallacy of Psychophysical
Metonymy. — Besides obscuring the true meaning of the true and
the false (as denoting the real and unreal respectively), this con-
fusion of the subjective with the objective senses of such words as
belief and judgment has had two consequences so important for the
understanding, or rather misunderstanding, of the realistic po-
sition, that I wish to consider them even at the cost of digressing
from our immediate issue.

.The first of these consequences of the 'verbal fallacy of psycho-

258 A THEORY OF TRUTH AND ERROR

physical metonymy,' is well illustrated by the principal Berkeleian
argument for idealism. That argument, as I have elsewhere
stated,1 consists in a seemingly valid syllogism in Barbara, whose
middle term 'idea' has, however, the same kind of ambiguity as
the term belief. The argument may be expressed as follows :
Ideas are incapable of existing apart from a mind. Physical ob-
jects in so far as they are perceived or known at all are certainly
'ideas.' All physical objects are, therefore, incapable of exist-
ing apart from a mind. Now this syllogism is formally valid,
and each of the premises is materially true. Its conclusion has
been accepted, albeit with reluctance and anger, by many genera-
tions of students of philosophy. The argument appears irrefut-
able until we notice that the middle term 'idea' is used in the
major premise to denote the act or process of perceiving, while
in the minor premise it is used to denote the object of that act,
i.e. the thing or content that is perceived. Each of these uses of
the term idea is familiar and justifiable in itself ; it is only when they
are identified with one another that the absurdity arises. Now a
more modern and widespread form of this fallacy consists not in a
play upon the word idea, but in a play upon the word 'experience.'
Experience has a good concrete flavor and is much affected by
empiricists. It is a term more comprehensive than idea, and less
suggestive of the over-intellectualism of the eighteenth century
psychology. It is regarded as such an ultimate sort of concept
as to need no definition. It is self-luminous, and everything is
to be defined in terms of it as constituting our ultimate universe
of discourse. Disliking to lay hands on so popular a fetish, I
must yet point out that, like belief, thought, and idea, the word
experience, which means first and primarily an act or process which
an organism performs or undergoes in becoming conscious of ob-
jects, has come by metonymy to be used also in a quite different
sense, denoting not only the process or state of being conscious
but the objects of which we are conscious. Now taken in this
latter sense, we may say that no objects exist outside the world

1 Program and First Platform of Six Realists, see Appendix.

THE BERKELEIAN FALLACY 259

of experience. For the only world that we can know or discuss
is the world that we experience. But to say this is very different
from saying that experience in the first sense, i.e. as a conscious
process, is a sine qua non of real objects. If we allow ourselves
to play fast and loose with the word, we get a theory which is in re-
ality thoroughgoing subjectivism but which has the appearance of
a sort of scientifically empirical compromise between realism and
subjectivism. If we ask of those who use the word in this double
sense for some statement as to what they have in mind when they
say "experience," we may be told that experience is that which
polarises itself into pairs of correlative and mutually reciprocal
opposites, such as subject and object, physical and psychical, in-
dividual consciousness and social consciousness, knower and
known; and that these terms and their relations are all within
experience, and constituted by it as its functions. When we ask for
some example of what this primordial stuff called experience is,
we shall doubtless be referred to tables, chairs, stars, mountains,
and such-like familiar objects. Suppose we ask further whether
these pieces of so-called 'experience' are not what common
sense means by 'things/ and whether the behavior of the things
which we experience does not clearly indicate that they antedate
and even condition our experience of them and are consequently
quite capable of existing independently of such experience. We
shall probably be told that the only things we can know are ex-
perienced things, that an experienced thing is in so far forth an
experience and that, as you cannot of course have an experience
which is not some sort of a conscious process, therefore objects '
cannot in any intelligible sense be believed or even imagined to
exist apart from consciousness. We may then bring up various
examples of 'external relations,' i.e. relations which are not neces-
sary to the existence of the terms related. We may cite the rela-
tion of pointing, where it is quite clear that the thing pointed at
does not in any way depend upon the fact that it stands in that
relation, and then ask whether the experiencing or cognizing of
an object may not be analogous to pointing ; from which it would

260 A THEORY OF TRUTH AND ERROR

follow that the thing experienced would not more depend on the
experiencing of it than a thing pointed at depends on the point-
ing at it. To which we shall be told that there is no analogy be-
tween pointing and experiencing and that it is essentially absurd
to distinguish between things known and the knowing of them.
The situation amounts to this : the constant use of the one word
experience to denote both an experiencing and an experienced
has produced in the mind of the idealist a curious delusion that
these things are connected, not merely verbally in his own mind
but materially in nature, and in such a way that the objects that we
experience can only exist at the moments when they are experi-
enced. Thus the consequence of the verbal fallacy of psycho-
physical metonymy has been to create an atmosphere and a ter-
minology that makes this experientialistic idealism seem not
merely true but axiomatic — indeed almost a matter of verbal
definition. Under such circumstances it is hard to get even an
understanding of what realism is. It appears indeed to these
experientialists so false as to be meaningless ; and those who de-
fend realism are actually charged with raising an issue that is
artificial. The only type of realism that possesses any meaning
in the eyes of the experientialists is the old Cartesian dualistic
theory of a system of objects wholly outside and beyond the world
of which we are conscious. But it is the distinguishing feature
of the new realism to maintain the independent existence not of
hypothetical objects which we do not experience, but of the very
objects that we do experience. And the comprehension of this
doctrine is impossible to any one who refuses to distinguish between
a thing that is experienced and the experiencing of that thing.

5. Second Consequence of .the Verbal Fallacy of Psychophysical
Metonymy. — The second consequence of the verbal fallacy of
psychophysical metonymy consists in a misconception of the
'laws of thought' and hence of the meaning and subject matter
of formal logic.

A law is a relation between things that is true or real, if not
at all moments, at least at more moments than one. These laws

THE DUPLICITY OF EXPERIENCE 261

differ in generality. Some laws hold of things only in so far as
they are extended ; other laws hold of all things in so far as they
are quantitative, irrespective of their possessing extent. Finally,
there are a few laws like the principles of non-contradiction,
and excluded middle and the dictum de omni, that hold of all things
alike, whatever their particular natures may be. These laws are
sufficiently important and distinctive to be the concern of a spe-
cial science — the science of logic. Now every law of things is a
law of thought in the sense that thought about those things will
be true or not according as it does or does not conform to their
laws. And in this sense the laws of chemistry and botany are laws
of thought. The laws studied in logic because of their absolute
generality are presupposed by the more concrete laws. And as
they hold good of every subject or topic to which thought can be
directed they must be taken account of by all thought regardless
of its more specific subject matter. It is only on account of this
universality that they have come by their misleading name of
"laws of thought." There is absolutely nothing psychological
about them. They are in themselves as independent of our con-
sciousness of them as the more concrete relations between
the elements of physics.1 And yet, it may be said, psychology
does enter into every treatise on logic to some extent and in
some way, and our conception of the purely objective and extra-
mental nature of the subject matter of logic would seem not
to allow for this. To which we may reply that psychology enters
in the same purely subsidiary way into the proper learning and
teaching of any other branch of knowledge. It is necessary not
only to learn the laws pertaining to any group of things but to
learn also the particular mental tendencies that aid or hinder
our mastery of those laws. Just as the astronomer, in order to
pursue effectively his study of the objective properties and laws
of the stars, must have some knowledge of the technique of the
telescope, the camera, and, finally, of the 'personal equation'

1 For a more extended demonstration of the non-mental nature of the laws of
logic, cf. Marvin, supra, 54 seq.

262 A THEORY OF TRUTH AND ERROR

or peculiar features of his own mentality which are likely to distort
his observations, so the logician, in order to understand and obey the
objective laws of logical implication, must have some knowledge of
the mental habits and tendencies which are likely to interfere with his
conforming to those laws. In short, the psychological study of
general types of fallacy is a necessary but purely subsidiary ad-
junct to logic, just as the psychological study of more special types
of fallacy is a necessary adjunct to the chemist or the statisti-
cian. Logic, like every science, has its special psychology and its
special pedagogy as parts of its technique. But to infer from this
that logic is a mental science would be as wrong as to infer that
astronomy was a branch of the science of optics.

6. Summary. — To sum up this first section of our discussion:
I have tried to show (1) that truth and error are respectively the
belief in what is real and the belief in what is unreal ; (2) that by
real, or true, is meant the totality of propositions comprising the
spatial and temporal system of interrelated 'events' or 'elemental
particulars/ together with all that may be presupposed or implied
by this system ; and that by unreal, or false, is meant the totality
of propositions, which are contradictories of the above, and
which comprise all the actual or possible objects of thought that
neither have a place in the spatio-temporal system nor are implied
by it ; (3) that the objection to identifying the true and the real
is based on the failure to realize that when true and false are used
as adjectives modifying belief, the word belief is used in the physi-
cal or objective sense of 'thing believed' rather than in the psy-
chical or subjective sense of 'believing'; (4) that this 'verbal
fallacy of psychophysical metonymy,' which consists in the con-
fused identification of the objective and subjective senses of such
words as belief, thought, idea, experience, has resulted not only
in the erroneous notion that the true is mental in its nature and
different from the real, but in the formulation of a proof of ideal-
ism which, though seemingly axiomatic, is in reality grossly equi-
vocal; (5) that as a further consequence of this fallacy of

LOGIC AND PSYCHOLOGY 263

metonymy there has arisen the false notion that logic is a branch
of psychology, and that the ' laws of thought ' relate to our think-
ing instead of to what we think about.

II

CAUSALITY AND CONSCIOUSNESS IN A WORLD OF PURE FACT

1. Space, Time, and Quality as the Ultimates of Factual Analysis.
— We have now to discover the meaning and place of con-
sciousness in the world of its objects, and to this end we must
examine again and more closely the nature of that world. First
of all, the world is a spatial system. Its groups of qualities are
exemplified in a three-dimensional reversible manifold of posi-
tions, each of which is outside of every other. But secondly,
this spatial system is not merely spatial, it is also temporal — each
position 'extends' infinitely into the past and future. This tem-
poral manifold, if for purposes of analysis we abstract from its
asymmetrical or irreversible character, can be regarded quite as
though it were a fourth dimension of space, and as each spatial
position or point extends infinitely in time, so does each temporal
position or instant extend infinitely in space. Looked at in this
way the existent world can be characterized as a four-dimensional
manifold of quality-groups, the units of which would be the quali-
ties actualized in any one place at any one moment. It would be
easy from this point of view to describe as functions of space, time,
and quality various less fundamental concepts. For example,

(1) One quality-group, one space, one time = event.

(2) Same qualities, different spaces, or different times = quali-
tative identity; similarity; species; class.

(3) Different qualities, same space, same time = numerical
identity; coinherence of attributes in one thing; 'isness.'

(4) Same qualities, in same space at continuously different
times = duration; rest. (Russell.)

(5) Same qualities, in same space at discontinuously different
times = succession, time interval.

264 A THEORY OF TRUTH AND ERROR

(6) Qualities partly same, partly different, same space, different
times = one thing undergoing change of state.

(7) Same or different qualities, same time, continuously differ-
ent spaces = one extended thing.

(8) Same or different qualities, same time, discontinuously
different spaces = plurality of extended things, distance.

(9) Same qualities, continuously different spaces, continuously
different times = moving thing. (Russell.)

(10) Qualities that change with change of space and time re-
lations = mere states; accidents.

(11) Qualities (if such there be) that remain unchanged
through all change of space and time relations = ultimate ele-
ments of quality; permanent substances.

Now if the reader will overlook the schematic and incomplete
character of this analysis, he will perhaps admit that it serves to
describe fairly the world of sensible objects. So far as I can see
there is no merely objective situation but what can be adequately
described as some function or complex of our three ultimates —
space, time, and quality. But there are two categories that we are
accustomed to recognize in our world which we do not find in the
system just portrayed. First, no place can be found in this sys-
tem for 'causality,' i.e. for a thing determining and being determined
by other things. And secondly, no place can be found for 'con-
sciousness,' i.e. for a thing cognizing and being cognized by other
things. The system before us is purely descriptive and purely
objective and contains no trace of the productive or the subjec-
tive functions. That both causality and consciousness are real
in some sense, none will deny. That an otherwise adequate de-
scription of the world which leaves one of them out should leave
out the other also, is, as I shall try to show, profoundly signifi-
cant of the relation between them. But before considering their
relation, we had best consider them separately.

2. The Antinomy of Causality — Substantist ' Thesis ' and Posi-
tivist 'Antithesis.' — We find as a matter of observation that there
are a number of uniform connections between the event-elements

THE CATEGORIES OF PURE FACT 265

composing our spatio-temporal world. Each of these event-ele-
ments, it will be remembered, was defined as a quality-group oc-
cupying some one space at some one time. In the simple
continua of space and time and in the dual continuum of space-
time or motion these events were all external to one another and
contingently related in the sense that there was no reason why any
event should not be related in any one of these continua to any
other event. And yet we find, with respect to any specific quality-
group, that there are two other specific quality-groups that respec-
tively precede and follow it, usually if not always ; e.g. fire usually
burns wood, water usually quenches fire. Now, as far as our space-
time system goes, all sequences should be determined by chance,
and the probability of wood-in-fire being followed by smoke-and-
ashes would be at least no greater than the probability against it.
In fact, as there are always fewer ways in which a complex can
happen than in which it can fail to happen, the probability that
any given sequence of events would regularly recur is indefinitely
small. And yet sequences do regularly recur, and indeed the more
we know of nature the more uniformities we discover. It is
clear then that in our analysis of the existing world we have so
far omitted to take account of a certain relation which is not
spatial, temporal, spatio-temporal, or qualitative. It is the rela-
tion of determining and being determined by, of cause and effect.
Each event-element has over and above its own qualities and its
own position in space and tune, something which implies or re-
fers to other events. It is both an agent and a patient of what is
not itself. Let us call this agent-patient property of a thing its
potentiality or power. The simplest examples as well as the most
useful of this category of power are to be found in connection with
the atoms, ether, and energy of physical science. The ideal atoms
or corpuscles of physics are things, all just alike, 'of little or no
extent, that preserve whatever properties they may have un-
changed from moment to moment of time and from point to point
in space. The primary function of the atom would seem to be the
power to give, receive, and maintain motion. The ideal ether of

266 A THEORY OF TRUTH AND ERROR

physics would be one permanent medium of great or infinite ex-
tent, the primary function of which would seem to be the power
to produce, receive, and transmit electro-magnetic oscillation.
Energy could perhaps be defined as that power in moving atoms
or ether which causes their motion to persist or at least to recur
unchanged through time. The great achievement of modern
science consists in the correlation of the various kinds of hetero-
geneous qualities and qualitative changes with the continuous
and homogeneous relations and changes of relation between these
perduring qualities or substances. The system of their space and
time relations, functions as a common denominator to which the
incommensurate qualities of matter and energy can all be reduced,
and so compared one with another. It is consequently very natural
that under such circumstances power should in some sense be
ascribed to these all-important elements, but there are two fun-
damentally opposed" methods of interpreting the nature of that
power. These two methods or attitudes I will call the posilivistic
and the substantistic.

The positivist recognizes the value of attributing causal power to
these substances when describing the actual world of qualities, but
he says their sole reality as powers is methodological. They are
'useful fictions' or 'shorthand formulae' by means of which we
can describe, predict, and control the routine of qualitative
changes in actual concrete objects. They are permanent pos-
sibilities of what is real, but they are not real themselves.
• They are, in short, merely subjective. To this the substantist replies
that it is absurd to relegate to the status of methodological fictions
the very powers in terms of whose activities the whole realm of qual-
ities can be expressed. It is rather the manifestations or appear-
ances of power which should be regarded as unreal or merely subjec-
tive, or as at best, mere 'epiphenomenal ' states of the interacting
substances. As to what further nature these substance-powers
have and as to whether they differ in kind, the substantists differ
from one another. Historically, there seem to be four main types
or schools of substantism, which I will briefly enumerate :

THE ANTINOMY OF CAUSALITY 267

(1) Materialists — Those who content themselves with postu-
lating the existence of power only in the kind of substances use-
ful in physical science, the atoms, ether, etc., of which we have
been speaking.

(2) Spiritualists — Those who postulate the existence of power
in certain obscure but perduring elements of self-consciousness
which thereby become 'souls' or 'egos/ i.e. powers, which mani-
fest their efficiencies primarily in the psychical and only second-
arily in the physical quality-groups.

(3) Dualists — Those who attribute power to both kinds of
substances — to those which interact in space to produce ' states '
of matter and to other entities which act upon those of the first
kind and perhaps upon one another to produce 'states' of mind
and of spirit.

(4) Energists — Those who reduce their hypostatization of
power to a minimum and merely affirm the existence of a force
or energy associated with each momentary quality-group such that
it tends regularly to produce a certain other quality-group as
its^ effect.

To all four of these schools of substantist doctrine, we may
imagine the positivist to reply in somewhat the following manner :
"Your hypostatized powers differ from one another only in that
some of them, such as the atomic, are useful methodologically,
and others, such as the psychic, are not. Whether useful or useless,
whether many and elaborately different, or whether reducible
to one type, in any case you can say nothing about their specific
nature, except in terms of their behavior or effects. The cause is
nothing but the permanent possibility of its effects. The meaning
of a power's own nature is nothing but the sum of its consequences.
The 'cash value' of any kind of potentiality is the totality of its
manifestations. Because it is methodologically useful to name a
thing in terms of what is to follow from it, we must not delude
ourselves into thinking that the potentiality of consequences
is anything objectively real. Causal power is not a subtle prop-
erty hidden behind the actual qualities of a thing; it is simply

268 A THEORY OF TRUTH AND ERROR

a convenient expression for our expectation that the future conse-
quences of a thing will resemble its past consequences. There is
absolutely nothing in any one thing that determines, necessitates,
or produces any other thing."

Apparently the only answer that the substantist can make to
this presentation of the positivist position is to reiterate his state-
ment that unless there were actually resident in each event some-
thing which determined or implied other events, the uniformities
of nature would be miraculous and we should have no right to
expect them. For to deny the objective reality of causal power
would mean a purely chance distribution of events in their rela-
tion to one another. And as we have seen, if mere chance held
good in the world, the probability of any sequence being regularly
repeated would be infinitely small.

The situation amounts to this: As long as the substantist
maintains that causal power or, more generally that the self -tran-
scending implication resident in events is not merely subjective,
his position is impregnable ; while as long as the positivist main-
tains that this potentiality or implicative reference of an object
can never be found among the qualities of that object, his position
is impregnable. Let us leave our two opponents to continue a
quarrel which, from the nature of the case, must be interminable,
and turn to the consideration of the second defect in our spatio-
temporal system of events — its failure to provide any place for
consciousness.

3. The Main Antinomy of Consciousness — Panhylist 'Antithe-
sis ' and Panpsychist ' Thesis. ' — There are two opposing ways of
meeting the problem before us — neither of them satisfactory.
One of them is called panpsychism, and the other I shall call
panhylism.

The panhylist declares that the only things that are actually
real are physical or objective things. What we call conscious-
ness or the psychical is for him not a true existent, but either
an epiphenomenon, a shadow or mirage that accompanies brain
processes, or else a mere possibility of the concomitant variation

SUBSTANTISM AND POSITIVISM 269

and 'togetherness' of events that are not together in any other
sense. The first form of panhylism is dualistic, the second monis-
tic. Their main agreement consists in a kind of psychophobia,
a hatred of consciousness and a determination to drive it out of
the world of real existence. The dualistic form of the doctrine
tries, as we have said, to accomplish this derealization of the psy-
chical by relegating it to the status of an epiphenomenon, or passive
correlate of the cerebral events. And yet, were it such, it could
neither sustain cognitive relations to the objects it appears to know,
nor causal relations with the stimuli, sensory and motor with which
it appears to interact. To say with this theory that the psychi-
cal is the 'other side' of the physical, or 'parallel' to it, is to use
categories which by their definition apply only to relations between
objects and not to the relation between objects and something
that is not an object. It is only because of its supposed methodo-
logical necessity for physiological psychology that the view is
tolerated at all. To accord any other than this epiphenome-
nal status to the psychical appears to violate the physical con-
tinuity and homogeneity of organic action and with it the doc-
trine of the conservation of energy.

But if the dualistic form of panhylism is bad, the monistic is
worse. For according to it the psychical is not even a helpless
epiphenomenon. It is simply a possibility that a living organism
possesses of varying with, or behaving towards, certain objects at a
distance from it in time and space. There are several varieties
of this doctrine, no one of which appears to me very intelligible.
The trouble with all of them is that they deny the existence of
that which is more certainly real than anything else, viz., my
awareness of objects. I have the experience, let us say, of per-
ceiving a chair. I am, however, forbidden by the monistic form
of panhylism to describe the experience in this way. I must not
say it is a case of 'my perceiving a chair.' I must only say it
is a case of 'chair.' But now my consciousness of chair cannot
be merely a case of chair, for there was and will be a case of
chair before and after my being conscious of chair.

270 A THEORY OF TRUTH AND ERROR

Obviously there is a difference between my consciousness of
objects and the mere objects, for my consciousness of them comes
and goes according to conditions in my brain. But they do not
depend on my brain processes. Moreover, if the consciousness
of objects is just objects, how shall I deal with a case in which you
perceive one thing occupying a given place in the spatio-temporal
series and I perceive another and contradictory grouping of
qualities in that same place and time? If two contradictory sets
of qualities could occupy the same space at the same time, we
should have no legitimate way of distinguishing between the real
and the unreal. In the effort to get rid of the subjective we
should have got rid of the objective.1 Perhaps the panhylist
at this point invokes the confused and vicious concept of
'true for me.' It is true for me, he will say, that the thing
is round and true for you that the thing is square. The
thing itself is both round and square. But even if we allowed
the panhylist to fill his now thoroughly Protagorean world with
round squares, and noisy silences, his notion of true for me and
false for you would still not avail, for in that world there could
be no 'me' and no 'you' for things to appear to, nor any meaning
to the word 'appear.' There would exist nothing but a stew and
welter of contradictions. In short, it is futile for panhylism to
deny the reality of the psychical and to treat consciousness either
with Huxley and Haeckel as an epiphenomenal shadow of physio-
logical processes or with Hume as a mere possibility of relativistic
and mutually incompatible fields of presentation.

A recent and very piquant form of monistic panhylism is the

JI regard the acute criticisms advanced by Professor A. O. Lovejoy (J. of Phil.,
Psychol., etc., 8, 589, seq.) as perfectly valid insofar as they bear against the
theory, apparently held by certain English realists, that hallucinatory objects exist
in the spatio-temporal system. So far as I can see the only way to escape (with-
out falling into either epistemological dualism or subjectivism) the ' ultra-realism '
or 'monistic panhylism' which Lovejoy attributes to the New Realists, and which
I have been attacking, ia to adopt the theory of the genesis of error and the non-
existent though objective subsistence of its objects, set forth in the third division of
this study under the title of the Epistemological Triangle.

OBJECTIONS TO PANHYLISM 271

identification of consciousness with behavior.1 The study of animal
psychology has shown the uselessness of trying to guess what may
be in an animal's consciousness, and the importance of concen-
trating on the study of the animal's behavior, which is something
actually observable. Moreover, the students of human psy-
chology have come to an increased realization of the importance
of studying the behavior of man, not merely as a means of in-
ferring what he is conscious of, but as something in itself pro-
foundly significant of his real nature and character. A study of
purely objective behavior does indeed in many cases afford a
better insight into the nature of human faculty and the means
of controlling and training it than could be yielded by any amount
of mere introspection. Not content with developing this new
and splendid branch of psychological inquiry as coordinate with
the study of consciousness, there have been some investigators
who, like Professor Singer, have allowed themselves to be so carried
away with enthusiasm for it that they have proposed actually
to identify consciousness itself with behavior. Against this latest
form of monistic panhylism the following objections seem to me
to hold valid : (1) Behavior is only one of many purely objective
processes. We can be conscious of behavior as of anything else.
But if behavior is itself consciousness, there seems nothing left
in terms of which we can define the consciousness of behavior.
In short, we have here the same sort of difficulty that confronts
the crude materialist who would identify consciousness with
motion. Motion has an actual qualitative nature of its own,
which is obviously different from the nature of consciousness.
(2) Behavior is always a movement or chain of movements in
space either of the organism as a whole or of something in the
organism, such as a neural current. As such it could at best only
be correlated with the consciousness of bodily movements, and
with what entered into them as their constituents. Now the
square root of minus one is not a bodily movement nor does it

1 (Cf. the paper entitled Mind as an Observable Object, by Professor E. A.
Singer, Jr., J. of Phil., Psychol., etc., 8, 180, seq.)

272 A THEORY OF TRUTH AND ERROR

enter into such movement in any intelligible sense. I cannot move
towards it or away from it. And the same is true of past events.
I do not see in what sense my consciousness of such things as the
life of Julius Caesar can be identified with any specific behavior
or movement of my body. (3) All that is visible or profitably
observable as behavior relates to movements with which it is
physiologically impossible for consciousness to be identified or
even directly correlated. For physiology teaches us that con-
sciousness depends upon, or is immediately and directly bound
up with, neural currents which are always intra-organic, if not
intra-cortical. Now what we observe as behavior in a man, a bird,
or a rat, is never the flow of neural currents, but only the gross
movements of the body and its members. These latter can by no
possibility be the physical correlates of the consciousness involved.
Still less can they be identical with it. (4) Finally, consciousness
does at each moment of a train of conscious behavior have for
its contents past incidents of the behavior that are no longer and
future incidents that are not yet. But it is obvious that with
respect to the behavior itself all its incidents are successive and
so outside one another in time, the past and the future never be-
ing present together. Let us note finally that most of these ar-
guments against identifying consciousness with behavior would
apply with equal force against any attempt to define conscious-
ness in terms of a physical movement, reaction, relation, or any
objective process whatsoever.

Let us now consider the opposite theory of the relation of con-
sciousness to the world of objects — the theory of panpsychism.1
The panpsychist might be defined as an idealist with a scientist's
conscience. He begins by recognizing the spatio-temporal order
of physical events. He takes particular account of the facts of
physiology. He observes that wherever and whenever certain

1 Panpsychism really connotes two distinct theories : (1) the positive view that
all matter has something psychical about it; (2) the negative view that all matter
is nothing but psychical. With the first of these views I am in entire sympathy.
It is only against the second or negative doctrine that I shall argue.

CONSCIOUSNESS AS BEHAVIOR 273

types of brain processes occur there is reported a consciousness
of events other than those brain processes. His own brain he
cannot observe, but his own consciousness he knows, and this
leads him to his great discovery, that the actual or 'in itself
reality of the brain process is the consciousness associated with it.
But if there is nothing except what is mental, why does everything
appear as physical? In short, why should the mind have a body
if there is no body for it to have ? To this question he answers by
saying that minds appear to one another as material ; that matter
is the form a mind takes when perceived externally by another
mind. My brain exists as such only as a state of your mind,
your brain exists as such only as a state of my mind. I will not
consider in detail the various forms which this doctrine has taken.
Leibniz, Schopenhauer, Clifford, Paulsen, Dr. Morton Prince,
Professor C. A. Strong and, just recently, Dr. Durant Drake,
have all presented theories which are essentially panpsychistic
in that they treat every material element of the world as in
itself mental, their materiality being nothing actual, but only
the guise or disguise in which they appear to one another. Pro-
fessor C. A. Strong has presented the view most plausibly in his
book, "Why the Mind Has a Body"; and so far as I know, he
is the only member of the school who recognizes the obligation to
answer the question why in a world in which nothing is material
everything should appear as material.1

I shall note briefly what appear to me to be the three principal
objections to panpsychism.

(1) The view offers no explanation of one mind's consciousness
of another mind as such. If every mind has to appear to another
mind under the form of matter, how happens it that we all of
us, and notably the pansychists, can think of, believe in, and talk
about other minds as minds ?

1 Despite the fact that this question, picturesquely formulated, makes the title
of Professor Strong's book, it is not in the book itself but in a paper published
later that we should look to find his answer. Cf. Archives de Psychologic, Nov. 1904.
And for my criticism of the argument there presented, cf . /. of Phil., Psychol., etc., a,

626.

T

274 A THEORY OF TRUTH AND ERROR

(2) The view discriminates arbitrarily between the forms and
qualities of the physical world. It denies the reality external
to the knowing mind of the colors, sounds, densities, and shapes
of the objects known, but it admits the independent and ex-
ternal reality of the space-characters of position and distance.
Its elements of mind-stuff exist and operate not only in time but
in a three-dimensional medium identifiable with the space of
material objects. Why should the panpsychist reduce known
matter and its qualities to the idealistic status of mere sensations
of the knower, and yet allow the spatial and temporal proper-
ties fundamentally constitutive of that matter to go scot free and
maintain a realistic status of independence of the knower ?

(3) Panpsychism is what Spaulding calls a ' self -refuting system,'
in that its conclusions contradict the premises from which they
are derived. For the panpj^schist at the beginning of his ar-
gument takes his stand firmly on the facts of physics and physi-
ology, thereby proudly differentiating himself from the ordinary
idealist. By appealing to these facts he demonstrates that the ob-
jects of which we are conscious and which seem to be real exter-
nally to us and independent of our knowing them, do in some sense
vary concomitantly and immediately with our cerebral processes.
They are, in short, only perceived by virtue of the ether and air
vibrations which they project upon the organism and which in their
turn produce nerve currents in the brain. On the basis of this he
asks us to admit that the objects immediately known are not really
external, but internal — identical with the cerebral processes
of the knower. Very reluctantly and with a sense of bewilder-
ment we may consent to admit this, violating our common sense
out of our respect for the physical and physiological evidence which he
has adduced. Then of a sudden the panpsychist makes a complete
volte face and cooly informs us that the ether and air vibrations
and the brain and its currents do not exist in themselves at all,
but are mere appearances in consciousness. My brain, in which
I have just consented to locate my entire objective world, turns
out to be merely a group of actual or possible sensations in

OBJECTIONS TO PANPSYCKISM 275

your mind. Now it is bad enough to have to locate my world of
objects in my brain considered as a thing actually existing and
receiving effects ; but it becomes preposterous to ask me to lo-
cate that world of my objects in what turns out to be not my own
brain at all but another person's state of mind. What shall it
profit a man to lose the whole world if he cannot gain, or retain,
even his own brain, — the one reality for the sake of which he had
made the sacrifice? Moreover, the whole physical and physio-
logical mechanism, on the validity of which the panpsychist ar-
gument was based, assumes a most anomalous position when once
the panpsychist conclusion is attained. For in place of a world
of bodies sending forth vibrations to one another, we must now
assume a world of naked consciousnesses which interact in an
utterly incomprehensible telepathic manner, the resulting content
of each being the unreal appearances of the others. The best
possible status that we could accord to physical processes in
such a world would be to describe them either as permanent
possibilities of sensation or as the epiphenomenal or shadow ac-
companiments of mysterious psychical interactions. Yet, once
more, it was only by assuming that these same permanent possi-
bilities or epiphenomenal shadows were not mere possibilities or
shadows at all, but a real material world, that the argument
could get under way. I do not wish to offend, but the whole
procedure of. the panpsychist in invoking physiological facts to
prove the unreality of the physical is comparable to that of the
man who climbs upon the shoulders of a friend only in order that
he may stamp upon his head.

Panhylism and panpsychism have proved themselves equally
futile in that they each degrade one aspect of things to a mere possi-
bility or shadow of the other. Yet, they are, after all, honest at-
tempts to explain the known in terms of the known ; and bad as
they are they are far better than the dualism and agnostic
monism which are their only rivals. An agnostic monism which
defines the physical and psychical as the miraculously parallel
attributes or manifestation of a substance or power whose nature

276 A THEORY OF TRUTH AND ERROR

is otherwise indefinable, solves no problems, either scientific or
metaphysical. What the substance is, why it is manifested as
^ other than what it is, and how its attributes are related as they
are — to these questions it makes no answer. To explain the two
realities that we partly know, it asks us to postulate a thing of
which we can know nothing. As for a dualism of the ordinary kind
which regards the world as made up of two ultimately separate
and perfectly heterogeneous entities, — objects and conscious-
ness— it not only offers no explanation of their interaction,
but by its very terms it makes of such interaction something that
is miraculous, if not impossible.

4. A Supplementary Antinomy of Consciousness. Are perceived
Objects inside or outside the Brain? — The world in which we
began our quest for consciousness, was a world which contained
nothing but events or quality-groups partly similar and partly
different, each group occupying its own spatio-temporal position
and related spatio-temporally to other groups continuous or
discontinuous with it. The objects of this world were all as
external to one another as are the points in space and the
moments hi time. Now what we denote as consciousness is
prima facie repugnant to such a system precisely because what-
ever else consciousness may be it is the condition by which ob-
jects at a distance from one another in space or time or both
may be in some sense 'present to' or 'in' or 'together with' one
another. To illustrate, I here and now am 'perceptually' or
'immediately' conscious of a noise that occurred two or three sec-
onds ago, and of a color two or three feet away ; and I am ' concep-
tually' or 'mediately' conscious of Julius Caesar in ancient
Rome and of the baby Emperor in Pekin. And here I wish (by
way of taking account of a supplementary antinomy of conscious-
ness) to call attention to four factors in this situation, two of
which make it necessary to regard these apprehended objects as
in me here and now, and two which make it necessary to regard
them as outside of me here and now.

(1) The objects of which I am conscious are in some sense di-

PANHYLISM AND PANSYCHISM 277

rectly effecting changes in me. They produce feelings of doubt as
to their appropriateness as illustrations ; they make me perceive
and think of other objects more or less related to them which would
scarcely have come to me except for them; they influence the
particular movements of my hand as I write of them.

(2) I have every reason to believe both from my experience of
dreams and imaginings, and also from what I learn from physio-
logical psychology that ' what objects ' I shall perceive or conceive
depends primarily and directly not upon the existence and activity
of those objects, but rather upon the existence and activity of my
brain. If the skull could be laid open and if by artificial stimula-
tion there could be produced in the brain and nervous system
the same processes — periodic waves, vibrations, stresses, or what
not — that are ordinarily produced by the energies that proceed
from extra-organic objects, then those objects would be appre-
hended quite irrespective of whether they existed or not. For
these two reasons it would certainly seem as though the things
of which I am conscious, but which appear as being at other
times and places, are nevertheless in me or in my nervous
system here and now. Yet (3) the things of which I am aware —
such things at least as I have chosen for the illustration — cannot
be within my brain because their immediate and incontrovertible
intent is otherwise. A color two or three feet away from my or-
ganism is not in my organism. The occurrence of a noise two
or three moments ago is not the occurrence of anything now at
this moment. The emperors of Rome and of China are not ner-
vous processes of a person in New York. To regard the things
as inside me, which by definition are outside of me, would be flatly
self-contradictory. (4) Moreover, all this aside, there is a second
reason that makes it impossible to identify the things known with
the processes occurring here and now by virtue of which I know
them. If you could look into my brain and see (and hear and
touch) everything that was actually there at a given moment,
you could not find there the things mentioned of which I was aware
at that moment. You might find, and, indeed, I believe you

278 A THEORY OF TRUTH AND ERROR

would find, visual and auditory simulacra of the things known,
'specifically qualified energies' streaming through the sensory
central and motor paths of the nervous system which would re-
semble more or less closely (according to the transparent or dis-
torting properties of the media through which they had passed)
the specific qualities of the extra-organic objects from which they
proceeded.1 But these cerebral effects, however closely they might
resemble the objects known, and in spite of the fact that they are
in some cases the effects of those objects, and in all cases the im-
mediate determinants of our knowing them, could not possibly
be the known objects themselves. If this followed from nothing
else it would follow from the fact that the parts of space and time
are mutually exclusive or external, and that the objects perceived
occupy spaces and times other than the spaces and times of the
brain process by which they are perceived. Here is certainly
a strange dilemma. Two factors in the situation seem to make it
necessary to regard the field of consciousness as 'in' the brain
and two other factors seem to make it necessary to regard that
field as not in the brain.

5. Consciousness and Causality. Hylopsychism as the Recipro-
cal Solution of the Antinomies. — In perplexity one naturally casts
about for analogies. Is there anything other than consciousness
that exemplifies the paradox of events in one space and time
being somehow in another space and tune? Yes. There is the
case of the causality relation. In considering that we found
ourselves confronted with the same absurd situation in which
an event in one place and time seemed to belong to or be de-
termined by a different event in another place and time.
There were, we may remember, two ways of viewing this causal-
ity parad6x : first the way of the substantist who insisted that
the self-transcending implications or potentialities of events

1 The evidence adduced by Holt in his criticism of the doctrine of specific energies
would seem to show that that very essential property of stimuli — their periodicity
— suffers little or no change in the passage from the merely physical to the physio-
logical medium. Cf. infra, 325 ff.

ARE OBJECTS INSIDE THE BRAIN? 279

were actually real in themselves over and above their specific
natures, and second the way of the positivist who insisted that
these potentialities were mere potentialities, nothing actually real
in themselves, but only our expectations that events would hap-
pen in the future as they had in the past. When the substan-
tist was asked to tell what the potentiality actually was he
could give no answer, and when the positivist was asked to
explain how a mere potentiality or subjective expectation could
explain nature's routines and uniformities which in a system
devoid of real causality and implication would be infinitely im-
probable, he could give no answer. And, as we have seen,
the quarrel between the panhylist and the panpsychist is as
interminable as that between the positivist and the substantist.
The only essential difference between the two quarrels would
seem to be that substantist and positivist refuted one another,
whereas panhylism and panpsychism seemed capable only of
refuting themselves. The truth is that the attempt to find a
place for causality in a world of pure facts leads to one kind of
antinomy and the attempt to find a place for consciousness in that
world leads to another. The last chance of attacking the prob-
lem successfully would seem to lie in the possibility that these
antinomies if matched against each other might, like the two
equally unintelligible halves of a picture puzzle, dovetail together
and reveal a clear and harmonious whole. Let us try this. The «c
substantist declared that a cause-effect implication resided in
each event, but the positivist has shown that such potentiality
could not be any new kind of quality. Suppose that this cause-
effect potentiality, which from the objective point of view can only
be defined indirectly as a possibility of other events, were in itself
and actually the CONSCIOUSNESS of those other events. Then causal
potentiality would have been made actual or real as the substan-
tist has proved it must be, while the positivist could still main-
tain that this potentiality was not any new and illegitimate type
of object. The essential claim of each party to the quarrel would
have been vindicated, but in such a way as not to conflict with

280 A THEORY OF TRUTH AND ERROR

the claim of the other. As to the second quarrel — that between
the panhylist and panpsychist — an equally satisfactory result
would have been reached. The panhylist could still maintain
that consciousness could be defined as the possibility of the objects
constituting its field and that as such it was not another real ob-
ject existing along with the objects revealed by it — while, on the
other hand, the panpsychist could claim that consciousness was
so fundamentally and immediately real that objects themselves
could be defined as permanent possibilities of perception. In
short, the positive claims of each party would have been satisfied,
and only the negative claims, which alone had brought about the
conflict, would have been rejected. Thus, the panhylist could no
longer say that consciousness was nothing but the possibility of
objects or nothing but an epiphenomenal correlate of the brain
process, for consciousness would have been made actually real
though in such a way as not to duplicate or interfere with the con-
tinuity of physiological processes. And the panpsychist, for his
part, could no longer say that physical objects were nothing but
appearances in consciousness, for a way would have been found
for the physical to appear in consciousness without prejudice to
its intrinsic reality as physical. Finally, as for the 'supple-
mentary antinomy' which turned on the reasons for and against
locating the objects of which we are conscious in the brain processes
which determine that consciousness, the matter would have been
cleared up in a manner equally satisfactory. fFor we can say that
with respect to their actual objective or physical being the events
that compose the world are as external to one another as the
spaces and times which they occupy, and in particular that the
events known are outside the brain events by which they are known ;
while with respect to their potential, subjective or psychical
being the events that compose the world are, in so far as they are
all causally related, internal to one another in the sense that any
or all might be present in each, and that in particular the events
known are 'inside' the brain events that know them, though
only as the latter's potentialities or implications. In short, the

WHAT CONSCIOUSNESS IS 281

relation between the brain-event and the object it knows would
be analogous to the relation of a word and its meaning. The
meaning is 'in' the word and yet the thing meant may be some-
thing wholly outside of and different from the word./
\ This is, I believe, the true theory of the nature of consciousness
and of its relation to the objective world which it reveals and in
which it abides, and because I believe so firmly in the truth of the
theory I wish to give it a name that will both distinguish it from
other solutions of the same problem and also show its peculiar
relation to those solutions. I will call it " Hylopsychism " after
the analogy of hylozoism — to indicate the especial synthesis
which it purports to effect between a hylistic or materialistic
account of things and a psychistic or (ontologically) idealistic
account. By hylopsychism then, I mean the theory that — The
potentiality of the physical is the actuality of the psychical and the
potentiality of the psychical is the actuality of the physical. Or, to
put it in the form of a definition of consciousness : Consciousness
is the potential or implicative presence of a thing at a space or time
in which that thing is not actually present. That there is much in
this formulation that is in need of further elucidation and proof
I am, of course, keenly aware. And there are several rather pal-
pable objections that must be considered. The arguments thus
far advanced have been mainly indirect — attempts to prove
what consciousness is by eliminating the various things which
it is not. That this indirect evidence needs to be supplemented
by evidence that is direct, I admit, and I believe there is an as-
tonishing amount of such evidence to be found by any one who is
willing to look for it. Some small part of it I shall try to bring out
in what follows.

6. The Three 'Directions' of a Potentiality. — Each event is (1)
an effect of earlier events, (2) a cause of later events, (3) in recip-
rocal interaction with the contemporaneous events externally
continuous with it in tri-dimensional space, which means that
each event faces in three ways — backward into the past, for-
ward into the future, and outward into space. If we were right

282 A THEORY OF TRUTH AND ERROR

in identifying consciousness with the cause-effect potentiality
or self-transcending implication functionally resident in events,
then we should expect to find the above-mentioned triplicity of
direction instanced in consciousness; and of course that is just
what we do find hi what are called memory, expectation, and ex-
ternal perception. Consciousness does at each indivisible instant
of its existence possess in greater or less degree a retrospect of the
past, a prospect of the future, and a perspective view of external
presence.

7. A Certain Difficulty in Terminology. — The foregoing analysis
seems to afford a suitable opportunity to explain a certain loose-
ness or ambiguity which the reader may have noticed in my use
of the words ' implication ' and 'potentiality.' I have used these
words as though they were synonyms, gliding unscrupulously
from one to the other as best suited my purpose. But this is, I
believe, more the fault of the terminology at my disposal than
of the thought itself. The fact is, 'potentiality' expresses one
aspect of my meaning and implication the other, and each is by
itself inadequate. There is no one recognized word for the three-
fold self-transcendent reference of events to their causes, effects,
and interacting contemporaries. The word 'reference' is itself
defective in that it seems to beg the notion of consciousness
which I am trying to define. 'Potentiality' is bad because it
is too narrow, meaning primarily simply the reference of a cause
to its future effect, and never, or hardly ever, the equally self-
transcending reference of effects to their causes. The word
'implication' suffers from the opposite defect to that of 'poten-
tiality,' being too broad. It can be used equally well to desig-
nate the reference of an event to its cause and to its effect, but
it is also used to designate the apprehended references of objects
to one another in a field of consciousness, and again to designate
the relation of the premises of an argument to its conclusion.
Now, needless to say, I do not mean to identify consciousness
with implication in either of these latter senses. A glass of water
can suggest to me the quenching of my thirst, without being

CONSCIOUSNESS AND POTENTIALITY 283

itself conscious of the use that I can put it to; and the major
and minor premises are as little conscious of the conclusion they
imply. In these cases the implicative relation holds between
complexes of events and not between the simple events themselves.
Water only implies thirst-quenching extrinsically as part of a
complex composed of instincts and memories; and again, the
premise or premises only imply their conclusion in virtue of their
membership in complexes that are more or less elaborate. These
logical implications of inference always obtain between relations,
aspects or complexes of event-elements and not concretely intrin-
sically and synthetically between the simple events themselves.
Because of these terminological difficulties I must ask the favor
of being allowed to continue to use (as an equivalent of the cogni-
tive function) the word 'implication' in the artificially narrow
sense in which it applies to the cause-effect relation, and the word
'potentiality' in an artificially broadened sense to denote the back-
ward reference of an event to its cause as well as the forward
reference of a cause to its effects.

8. The Three Levels of Potentiality. — Hylozoism is the theory
that all matter is instinct with life. By hylopsychism I wish
to denote the theory that all matter is instinct with something
of the cognitive function; that every objective event has that
self-transcending implication of other events which when it
occurs on the scale that it does in our brain processes we call
consciousness. Now the reader will naturally take fright at
the idea of postulating anything like a human consciousness in
the movements of dead matter. And I hasten to say that
the theory that I am advocating will in no way obliterate the
pragmatic differences between the mechanical, the physiological,
and the mental. Science has fought hard and long to free her
explanations from the incubus of teleology and to gain the right
to treat nature as a spatio-temporal mechanism ; and there would
be a strong presumption against any theory which would con-
flict with the depersonalized austerities of physical law. And
yet along with the conquering advance of the mechanistic idea

284 A THEORY OF TRUTH AND ERROR

has come an increasing sense of the continuity of all natural pro-
cesses— from the blind movements of an atom to the far-seeing
actions of man. Our enterprise must be to find a place for teleology
in nature, not as a substitute for spatio-temporal change, but ra-
ther as a special and higher type of such change.

There are in nature at least three fundamentally distinct types
of causal process, the mechanical, the vital, and the sensory. In
the mechanical process, the dominant characteristic is spatio-
temporal change. The position of a body at a given moment
implies and is implied by its own or another body's position at
another moment, and its 'consciousness' would be -of that and
only that. It would have a purely spatio-temporal content. In
the vital process, however, the dominant characteristic is change
of organic form and chemical pattern rather than mere change
of place. A living organism at each moment of its life contains
as its potentiality or implication organic forms that it has had
and will have, together with the potentiality or capacity of as-
similating or adapting to the chemical pattern of its own tissues
the appropriate matter which as food may come within its system.
Its potentiality expresses itself in inheriting, developing, and
assimilating; and its 'consciousness* would have for its content
organic forms or chemical patterns which it had undergone and
was about to undergo.

In the sensory process which constitutes our consciousness the
implications are neither of mere notion nor of inherited metabo-
lisms, but of specific qualities of the objects outside our organism
and causally related to it. That is to say, the nervous system grows
up within the organism as an apparatus for transmitting energy-
forms from one part of the organism to another and so from the
things outside to the brain inside. Most of the cerebral energy
thus produced is continuously passing over into motor currents
initiating behavior, but an infinitesimal part is dammed up,
or stored in the cortex. The continuous currents contain the im-
mediate moment to moment consciousness of sensations and
feelings, while the part that is stored up as potential energy has

CONSCIOUSNESS AND LIFE 285

for its implications at each moment the system of extra-organic
objects occupying other times and places ; thus it is the principal
function of the brain to receive and retain in something of their
separate specificities the energies that have come from distant
objects; and it is the self -transcending implications of these
brain-states that constitute our consciousness of the spatio-tem-
poral world in which we live.

We might, indeed, compare the sensory mechanism of the ner-
vous system with the merely vital mechanism of protoplasm in
general somewhat as follows: Protoplasm is an apparatus for
ingesting, digesting, reproducing, and excreting matter and the
chemical energies associated with matter. The nervous sys-
tem, on the other hand, takes its energy neat. And in perceiving,
remembering, imagining, and reacting it is respectively ingesting,
digesting, reproducing, and excreting those free energies dissociated
from matter which in the form of vibrations of various kinds have pro-
ceeded from distant objects through the sensory channels to the brain,
where they constitute by their implications a consciousness of those ob-
jects and make possible an intelligent and purposive adjustment to an
environment [extending in time and space immeasurably beyond the
field of mere chemical and mechanical contacts. All this, of course,
is very light handling of important and intricate matters. But it
may suffice as a hint of how the theory of hylopsychism could
meet the criticism that by its identifying consciousness with
causal implication it had unduly minimised the differences between
mechanical, vital, and rational processes. The three grades of
potentiality .here discussed retain all their characteristic distinctions.
We have neither anthropomorphized the laws of physics nor re-
duced to a blind and mechanical process the activities of the
human spirit.

9. Summary. — The main purpose of this second division of our
discussion has now been accomplished. Our definitions of truth and
error as the consciousness of the real and the unreal forced upon us
an ontological excursus or metaphysical inquiry into the nature
of mind and its relation to the world of its objects. The analysis

286 A THEORY OF TRUTH AND ERROR

of the existing world in terms of space, time, and quality brought us
face -to face with the antinomies of causality and consciousness.
As a joint solution of the two seemingly insoluble problems pre-
sented by those antinomies the theory of hylopsychism was pro-
pounded. The last sections of the chapter have been devoted
to an exposition and defense of that theory and to a brief account
of its bearing on the psychophysical and the biophysical prob-
lems. We are now free to return to the more directly epistemo-
logical question of the nature of truth and error.

Ill

THE GENESIS OF TRUTH AND ERROR

1. The Epistemological Triangle. — Physical objects send forth
waves of energy in various directions and of various kinds, but all
in some measure characteristic of the objects from which they
proceed. These energies impinge upon the organism, and the sen-
sory end-organs and the nerve fibers then transmit to the brain the
kinds of energy to which they are severally adjusted or attuned.
The final effect is the resultant of these sensory energies modified
by the reaction of the brain. This complex cerebral state is some-
thing quite physical and objective — as much so as the extra-organic
object which is its partial cause. It is a natural event with its own
qualities and its own position in the space and time order. As
such it possesses the threefold self-transcending implication or
cause-effect potentiality which we have seen to be the property of
every natural event, and which we have declared to constitute in
itself that event's consciousness of other events. The cerebral
state, whether initiated from within the organism, as in spontane-
ous thought and in hallucination, or whether initiated from with-
out as in perception, will be conscious of such objects as it implies
or of which it is the potentiality. What will these implicates
or objects be? My answer is that they will consist of the events
which would most simply have caused the cerebral state and of the

HOW THE BRAIN DIGESTS ENERGIES 287

events which the latter would produce as effects if it acted alone and
uninterfered with. Now we know that if we single out some one
event and inquire as to its cause, that we shall find a plurality
of possible antecedents, any one of which if it had not been coun-
teracted would have produced it. This is the principle of "the
plurality of causes." It follows from this that the implicate, or
conscious object of any brain state may be, but need not be, an
event which actually exists. When the implied possible cause
actually exists, then there will be consciousness of a reality which,
as we have seen, constitutes true knowledge or truth ; when, on the
other hand, the cerebral implicate which is the simplest or most
natural of the possible causes happens not to have been the actual
cause, or happens not to exist, then we shall have apprehension
of what is unreal, which is false knowledge, or error. The brain
event is the 'knower' and what it implies is the 'known.' Perry1
has amply shown that the implied is not dependent on the implier,
and hence the fact that the brain event implies its actual or possi-
ble cause as the object in consciousness does not mean that the
latter in any way depends on the former. Even when we consider
the case of the forward-facing, or prospective implication, the ob-
ject will not of necessity depend upon the knowing (or expecting)
it. For just as there may be more than one cause capable of pro-
ducing an effect, so there may be more than one effect produced
by a given cause. And because an earlier event is the poten-
tiality of a later event, it does not follow that it will inevitably
produce it. This statement, however, may seem to require some
defense. It may be said that if A is truly a cause of B then when-
ever A exists B must exist. Now, of course, the word cause can
be defined to mean whatever events the occurrence of which univer-
sally determine the existence of a later event called the effect.
And this is, I admit, the ordinary conception of cause in philoso-
phy. But it is not the common sense conception of cause. Ac-
cording to common sense the cause is not something that necessarily
produces its effect, but only something that tends to produce its

1 Cf. supra, 112.

288 A THEORY OF TRUTH AND ERROR

effect. The fact that the strict or philosophic conception of cause
does not allow for and even renders impossible and meaningless
the notion of the counteraction of a cause is in itself sufficient to
decide against that conception. If A could only be called the
cause of B when it was the universal antecedent of B, then it
could by definition never be counteracted. And yet, we often
speak of a cause being prevented from realizing its effect ; and I
shall therefore accept the lesser meaning for the word cause as
that which tends to produce and if unhindered would produce its
effect. Taking the matter in this way we can see that the impli-
cates or objects of an existing brain-state, whether those im-
plicates are its causes, its effects, or the contemporaneous and
reciprocal combination of the two, do not themselves certainly
exist, but only probably. Or to state the same truth in another
way : The consciousness that an event has happened or will happen
in the past or future carries with it not the certainty, but only the
probability that the event has actually happened or will actually
happen. But while the probable truth of the past event which
we are conscious of as having occurred is not in any sense depend-
ent upon our will, yet the probability of the actual occurrence
of a future event which we expect does often depend on the amount
of voluntary effort expended by ourselves or others. That is to
say, a potentiality will bring about its own future actualization
only when unimpeded. If there are other rival tendencies at
work, it will be necessary to counteract them by use of the will.
Perhaps I can best sum up this account of the genesis of the
consciousness of real and unreal objects by symbolizing the three
elements involved. We have (1) the actually existing external
object which I will call Oe; (2) the cerebral state itself which
may be denoted Oc; (3) the object perceived or apprehended,
Op. In the simplest case, Oe will be the cause of Oc and in
every case Op will be the implicate of Oc. These are the three
corners of the epistemological triangle. We might compare them
in their relations to a luminous object, Oe, its impress upon the
surface of a mirror, Oc, and the virtual image seen behind or

THE EPISTEMOLOGICAL TRIANGLE 289

through the mirror, Op. Now, sometimes the virtual image
may be identical with the luminous object, both in quality and
in position, as when, for example, a luminous object is placed be-
hind a mirror and then by means of a larger secondary mirror
with an appropriate curvature of surface behind the observer the
light from the object is reflected back into the surface of the first
mirror. A virtual image of the luminous object will then appear
having the same qualities and the same position as the object itself.
In the same way, the object Op revealed or implied in the brain-
state Oc may be exactly identical in position and quality with
the actually existing external object, Oe, or on the other hand
it may not. In the former case the object is apprehended as it
is, in the latter case as it is not.1

2. The Two Kinds of Truth and Error. — If truth and error are
generated in the manner symbolized in the ' epistemological tri-
angle,' it follows that there will be two ways in which each may
arise. In the case of truth, in which the real event Oe is
identical with the perceived event Op, the result may be due (1)
to the fact that the medium through which the energy has been
carried from the external object to the brain has not altered the
character of that energy, in which case the cerebral event Oc
will be the same in quality as Oe and the local and temporal
signs of the former will be of such a kind as to imply the real
time and place of Oe; or (2) the medium between Oe and Oc
may have qualitatively, temporally, and spatially distorted or
modified the energy proceeding from Oe, but the brain through
inherited capacities or through memory-traces will have neu-
tralized and corrected this distortion so as to make the ulti-
mate determining brain-state Oc so congruent spatio-tempo-

1 The reader must guard against a possible misinterpretation of the simile of the
triangle which I have used to illustrate the process of perception. No two vertices
of a triangle are ever numerically identical, whereas we have seen that in so far as
perception is true the perceived object and the real object are identical. Oc is the
effect-function of Oe, and Op is the implicate-function of Oc. And just as any given
if1 root of art may or may not happen to be a itself, so the implicate Op of the
effect Oc may or may not happen to be the actual cause Oe.
u

290 A THEORY OF TRUTH AND ERROR

rally and qualitatively with Oe as to make it implicative of
it. This second kind of truth is, I think, the more frequent type,
especially hi the case of the perception of external objects. In
vision, for example, the retinal image is notably different from
the real object, in being dualized, in being inverted, and in being
otherwise distorted, and yet whether instinctively (as the na-
tivists would hold) or by past experience (as the associationists
would say) the brain succeeds in neutralizing these distortions
and giving a perceived object which is often more truly identical
with the real object in all its attributes than a single effect even
if wholly undistorted could have revealed. Take for example
the case of the square table perceived as square. The retinal
images are rhomboidal and the first cerebral effects are probably
proportionately uncharacteristic of the external cause, but, re-
peated experience has served to generate a condition in the brain
(purely physical but otherwise analogous to the 'apperception
mass') made up of traces of the various visual impresses and the
motor attitudes excited by them, which complex condition is
implicative of the true form and position of the table. The cere-
bral implication may in short be either of the direct immediate
sensory type or of the indirect mediate and apperceptional type.
Probably the only case in which the first or sensory kind of truth
occurs is in what is called the ' consciousness of our own states ' ;
that is, the consciousness at each moment of the brain processes
and implications of the just preceding moment. In this way
and in this way only can we be conscious of consciousness or
self-conscious. The brain-state of a given moment is never
conscious of itself as object, but it can be conscious not only of
extra-organic objects, but also of the brain-states just preced-
ing it. And in this intra-organic consciousness, where the self-
transcending implication 'reaches' only to the next moment,
there would seem to be no chance for error. In other words, the
general impression that we can be more certain of our own thoughts
and feelings than of anything else would seem to be well-founded.
Before going on to consider the corresponding types of error,

THE TWO KINDS OF ERROR 291

it would be well to consider briefly the difference between an in-
ferred or conceived object and one that is perceived or directly
apprehended. I believe that the distinguishing characteristic
of indirect or conceptual consciousness is the spatial and temporal
discontinuity of the apprehended object and the organism. When
we perceive an object we are conscious not merely of it but of the
space and time which intervenes between it and ourselves. This
brings with it a motor attitude and a tendency to immediately
react, which constitutes our reality-feeling. In the case of the
conceived imagined or inferred object there is no accompanying
consciousness of the intervening space and time connecting it
with our body, and consequently no such definite immediate ma-
chinery of motor adjustment is set up.1

Let us now consider the two kinds of error. We can deal very
briefly with them. They both result, as we have seen, from a
distortion of the real object in producing its effect on the brain.
The distortion may be (1) physical or peripherally physiological,
in which case we have the so-called sensory illusions, or (2) it
may be central, due to the cerebral apperception mass, in which
case we have the error of inference. Of course we may have both
kinds of error together. And moreover, between the two kinds
there is no hard and fast distinction. You can say if you like that
"the senses never lie" and that even the so-called perceptual
illusions are illusions of inference. But this amounts to nothing
because it will immediately be necessary to add that the element
of inference is present in all perception, and that a sensation that
contains no content beyond the moment, or no element of self-
transcendency and no tendency to action, is an unreal figment
of the psychologist's imagination. The distinction between per-
ception and inference is a valuable one, but not of such a charac-
ter as to rob perception of its own kind of self-transcending ref-
erence. When a child clutches at and misses the stick which he
perceives as bent in the water, he does not infer from an internal

1 Cf. J. McK. Cattell, in Essays Philosophical and Psychological in Honor of
William James, 569.

292 A THEORY OF TRUTH AND ERROR

sensation or image of a bent stick to a bent stick existing outside
him. But rather, the image of the bent stick which is cast upon
his retina (as it would be cast upon a photographic plate) pro-
duces a purely physical brain-state which directly 'implies,' or
has for its 'meaning' or 'potentiality' an external bent stick. The
implicate of the brain-state, however, happens in this case not to
exist, and therein consists the error. The other type of error re-
sults when the deliberate inference made from a correctly perceived
object to another object not perceived,'is incorrect, in that the object
normally implied by the situation happens not to exist.

The two kinds of truth and error which we have been consider-
ing might be summed up and classified with reference to the place
in which the stimulus is distorted or corrected as follows :

(1) Absence of physical or peripheral physiological distortion
combined with absence of cerebral distortion = immediate or
sensory truth, as exemplified in feeling and introspection.

(2) Physical or peripheral physiological distortion, corrected
by cerebral and apperceptive reaction = mediated truth, as
exemplified perceptually in the apprehension of a square surface
as square, and conceptually or inferentially in the knowledge that
a thing is not what it appears to be.

(3) Physical or peripheral physiological distortion without any
corrective counter-distortion by the brain = immediate or sen-
sory errors and illusions and, when initiated from within the or-
ganism, hallucinations.

(4) Cerebral or apperceptive distortion combined with absence
ofphysical or peripheral physiological distortion = mediate errors
of conception and inference, which when persistent, constitute the
delusions of insanity.

3. Attention and Belief. — Why is the brain not conscious at each
moment of all the implicates or potentialities of the sensory-motor
currents and traces of currents which it contains ? or why is the
mind not always aware of its entire stock of knowledge ? And,
again, why are we always more aware, or more keenly conscious of
certain objects than of others? I believe that our conception of

THE TWO KINDS OF ERROR 293

consciousness makes possible a comparatively simple answer to
these questions. The motions and stresses in the cortex in so far
as they are in interaction must form a single system and set up a
condition analogous to that of a single body acted upon by many di-
verse and opposing forces. A body in such a situation has at each
moment a dominant tendency to move in the direction of the re-
sultant of the forces acting upon it. The intensity of this dominant
tendency or controlling bias will vary with two factors: (1) the ab-
solute total intensities of the forces; (2) the degree of unity of direc-
tion of the forces relatively to one another. Now we shall not
need to commit ourselves to any elaborate system of psycho-
dynamics to recognize that attention is 'absorbed,' or 'drawn/
in a given direction not only by the intensity of sensation,
but also by the extent to which there is absence of rivalry. There is
both an absolute and a relativistic factor in the situation. Now the
objects that we are conscious of at any given moment will be the
implicates of the resultant of cortical forces acting together. And
just as a body cannot move in both a northerly and southerly
direction at once, so for the same reason it will be necessary for some
of the cortical tendencies to be absolutely counteracted, or eclipsed
by the others. This would, I think, explain why a large part of
what is 'in the mind' must be extruded absolutely from the con-
sciousness of any one moment. But while one cortical tendency
may completely suppress another, it may also only fuse with it,
dominating it not absolutely, but only partially, and then we
should have the implicate of the dominant tendency as the cen-
tral and most prominent object in the field of attention with im-
plicates of the lesser but not absolutely counteracted tendencies
figuring as objects in the more marginal parts of that field. And
from this it would follow that except in unified and systematized
complexes, the intensity of attention would vary inversely with
its extensity. The mechanical analogy of the forces acting upon
a body is imperfect, mainly because it applies merely to movement in
tridimensional space, whereas in the field of cortical forces we
have not merely tendencies to movement, but all sorts of tenden-

294 A THEORY OF TRUTH AND ERROR

cies correlated \vith the higher derivatives of space with regard
to time. And the field of implicates or objects of consciousness
constitutes as it were a multidimensional manifold incomparably
richer than the mere system of space and time relations.

And now as to belief: it seems probable that the primary con-
dition of consciousness is a condition of acceptance of cerebral
implicates or conscious contents at their face value as real and as
bases for action. Disbelief and doubt are sophisticated or sec-
ondary attitudes which we take towards a content only when
it is contradicted by another content or by the system as a whole.
A person believes a thing or accepts it as real and tends to act on
it just in so far as there is no content present which suggests ac-
tually or hypothetically a counteraction. In hypnotism we seem
to have a case in which the ordinary associations are cut off and
there is nothing to inhibit action in accordance with what is sug-
gested. The mind thus relapses to its primary naive attitude
of belief in the reality of all its objects. But the field of belief
at any one moment will normally be much narrower than the field
of attention at that moment, for the same reason as that which
makes it possible for a body to tend to move in several directions
so long as they are not actually opposite, but prevents it from ac-
tually moving in more than one direction. So we may entertain
and compare a number of contradictory proposals without con-
tradiction, but we could not believe them all for the action based
on any one of them would contradict or preclude action based on
any other.

4. The Material Fallacy of Psychophysical Metonymy. — In the
first division of our essay we discussed the fallacy of identify-
ing the thing experienced with the experiencing of it, simply on
the ground that by metonymy, the same word 'experience' was
used to denote both the psychical act and the physical object of
that act. And we found that the two consequences of that fallacy
were first, a false but seemingly axiomatic proof of idealism and
second, a false conception of the nature of logic and the laws of
thought. Now there is a somewhat similar fallacy with similar

ATTENTION AND BELIEF 295

consequences which is based upon a misconception of the relation
of consciousness to its objects. ^ But this fallacy is not verbal, but
material. We can state it briefly as follows : Which objects I
shall at any moment be conscious of depends upon my own processes
and states, therefore the objects of which I am conscious depend
upon my own states and consequently cannot be real apart from them.
This argument, like the one considered before would, if it proved
anything, prove too much, for it would mean that other minds,
as well as physical objects, would depend upon being known.
But we may waive this point and content ourselves with a direct
refutation of the argument. It is true that 'what objects' or
'which objects' I am conscious of depends upon my cerebral
processes, but this does not imply a causal relation of depend-
ence between those objects and my cerebral processes. There
is indeed precisely the same concomitant variation between the
movements of my finger in pointing and the objects pointed at.
No one will deny that 'which objects' I point at will depend on
how and where I move my finger. But no one will affirm that
therefore the things pointed at depend on my pointing and cannot
exist apart from it. In the same way, 'which objects' I write
about will depend on the words I use, but no one will say that there-
fore those objects depend upon their own names. Now the re-
lation of the cerebral state as implier to the apprehended object
as implicate, with which relation we identified consciousness,
is in this matter of concomitant variation and dependence pre-
cisely analogous to the. relation of the pointing finger to the object
pointed at or the relation of a name to the object of which it is the
name. The relation or act is in each case selective and not cre-
ative or constitutive in any manner or to any degree. In short,
we cannot infer from the dependence of 'which objects' to a de-
pendence of 'the objects.'

The first consequence of the material fallacy of psychophysi-
cal metonymy is of course the one implied in the exposition just
given. It constitutes the principal proof of idealism; and as
an argument it seems less axiomatic, but more solidly empirical

296 A THEORY OF TRUTH AND ERROR

and convincing than the verbal fallacy of the same name. The
subjectivist presses his eye-ball and sees a chair that moves where
formerly he saw a chair that was stationary. This should suffice
to convince him that how you see a thing depends on how you see
it ; but not content with this modest inference he jumps to the
conclusion that the things which he sees depend on his seeing
them, — which is different. And, as we have said, he never stops
to consider that the other minds in whose independent reality he
believes are relative in just the same sense to the processes by
virtue of which he thinks of them.

The second consequence of this fallacy of metonymy consists in a
far-reaching and widely prevalent misconception of the nature of the
thinking process, and especially of the j udgment process. Judgment
is the utterance or expression of a belief, (tentative or certain) in the
reality of an identity-complex, or proposition. An utterance is an
act that takes a certain amount of time. The symbols of its
expression have to come in a certain sequence, e.g. subject, copula,
predicate. The judgment is often the outcome of many other acts,
and usually involves effort and activity of various kinds on the
part of the one who makes the judgment. Now all this belongs
to the biography of the one who judges. The fallacy in question
reads this chapter of the thinker's biography into the biography of
the object of his thought. We start out with the judgment, A is B ;
finding this untrue, we make a new judgment, A is C ; perhaps this
leads us to believe that A is not only C, but is also D and E. All
of which may be very interesting and important for us, and may
constitute a profound alteration and enrichment of our natures ;
but it is not interesting and important for the object, and it does
not constitute any alteration or enrichment in the nature of the
object. It is our privilege to learn, sometimes suddenly and
easily, and sometimes with much time and effort, the nature of
the real things about us ; and it is our misfortune that sometimes
we fail to discover reality and succeed in discovering only its con-
tradictory or shadow, i.e. the unreal and false. But all this is
neither a privilege nor a misfortune for the reality that we do or

THE MATERIAL FALLACY OF METONYMY 297

do not discover, except of course in the quite indirect sense that
our judgments may lead to behavior which may later effect the
objects discovered. Our judgments can remain unchanged even
when they refer to objects as changing as fire, or they can change
and evolve even when their objects are as changeless as the Sphinx
or the square root of minus one. And yet it is often assumed that
the world can be proved to be dynamic and changeful because our
judging process is that. There is no inference possible from the
changing or static nature of our judgments to the changing or
static nature of what is judged about. We do not break up the
living flux of reality when we conceive of its various phases and
aspects. Conception does not dirempt reality, and judgment
is not the process of pasting its pieces together. Conception
and judgment are, like all cognitions, not constitutive, but se-
lective. To conceive is to apprehend simply one quality or group
of qualities, and to judge is to apprehend an identity-relation
between different qualities or groups of qualities.1

5. Degrees of Truth and Error and the Fallacy of Internal Rela-
tions. — All or almost all of our cognitions are partly true and partly
false. Expressed in our own terminology, the brain states at any
moment have, for their self -transcending implicates, objects some
of which are real and some of which are unreal or merely sub-
j ective.2 The cause of this is, as we have seen, the fact that our cere-

a more systematic demonstration of this theory that judgment (like all
other forms of consciousness) is purely selective and never constitutive cf. my dis-
cussion of Dr. Schiller's Humanism in "May a Realist be a Pragmatist" /. of
Phil., Psychol., etc., 6, 565-6. With reference to the manner in which ' anti-intel-
lectualism ' by neglecting this view confuses the relations between the symbols used
in making judgments with the prepositional content asserted by the judgment c/.
/. of Phil., Psychol, etc., 7, 153-4.

* Anything in so far as it figures as an object in consciousness may be called ' sub-
jective ' in the broad sense ; but the word is more commonly used in the narrow
sense of ' merely subjective ' to denote objects of consciousness which are unreal or
non-existent, such as the events in a dream or the content of a hallucination. There
is nothing that is of necessity mental or psychical about these subjective objects.
But the fact that as non-existent or unreal subsistents they have no proper place
of their own in the spatio-temporal system and hence act only on and through the
mind that knows them led to their being treated as mental in the same sense as de-

298 A THEORY OF TRUTH AND ERROR

bral states are in part, and only in part, the effects of the objects
external to them and known by them. The physical medium
and the brain itself is always partly responsible for the cerebral
state. And the aberration thus effected is probably never en-
tirely corrected. * The source of error in other words is due to the
plurality of causes and to the counteraction of effects. Now the
understanding of the source of error at once suggests the means,
not indeed for its complete cure, but for an indefinitely progres-
sive amelioration. The more effects we have of things, the less
ambiguity there is in their joint implication ; and though each one
may have been partly modified, yet these modifications may be
neutralized by one another. A great number of pretty bad photo-
graphs of a person might suffice better to give us an accurate idea
of his appearance than one fairly good photograph. The totality
of a thing's effects, even though each had been interfered with,
would not, indeed, be themselves identical with the thing, but
they would be exactly and adequately implicative of it. And
the more effects we get of a thing, the more nearly adequate is
our knowledge of it. If the stick partly immersed in water could
effect us only through our eyes, and from only one point of view, we
should never discover the error of regarding it as bent. But it
effects us from many points of view, and through touch as well
as vision, and thus the error of immediate perception is discovered
and rendered innocuous. In this way we can test our knowledge
and attain ever more and more probability of truth.

From this standpoint it might be well to consider errors of
quality as distinguished from errors of position or numerical identity.
We may perceive a thing in the right place and time, but the quali-

sires and volitions are mental. This natural confusion was the first stage in the
evolution of subjectivism. It reduced all non-existent objects to the realm of mind
and led naturally to the view that all objects in so far as they were known thereby
became states of the knower. The truth is simply this : Real objects have two sets
of consequences. They effect other objects continuous with them in the world of
space and time and they also affect whoever knows them. Unreal objects, how-
ever, have no effects or consequences of any kind except upon the persona in whose
consciousness they figure.

DEGREES OF TRUTH AND ERROR 299

ties which we perceive it to have may never occur in it. They
may be qualities such as are only correlated with or produced by
energies peculiar to our nervous system. There would be real
error involved in such a situation, but it would not of necessity
result in harm to us. As long as the qualities really present in
the objects varied in one-to-one correspondence with the quali-
ties apprehended, we should be well enough off for practical pur-
poses. I may here remark that this may be our plight with re-
gard to the secondary qualities (though not with regard to the
primary qualities whose reality is vindicated indirectly as well as
perceptually). The problem of the external reality of the sec-
ondary qualities is neither solved nor intrinsically insoluble. The
way and the only way we could settle the matter would be to
gain exact knowledge of the primary energies occurring in, or on
the surface of, the external bodies, and an equally exact knowledge
of the primary energies in the appropriate sensory cerebral tracts,
and then compare them. In so far as they were the same or differ-
ent, we should be certain that the secondary qualities of things
were or were not as we perceived them.

Now if we allowed the fact that every object of consciousness
was partly true and partly false to make us believe that there-
fore the real world might be in every respect different from what
we perceive, we should commit the fallacy of internal relations.
For that fallacy consists in the assumption that the nature of the
parts of a complex depends upon the nature of the whole complex,
and consequently, that knowledge of merely a part of the truth
must as such be false. Spaulding has refuted this fallacy in his
defense of analysis. We do not have to have all truth in order to
have some truth. I can misapprehend some qualities of an object
without misapprehending its other qualities. And because I
erred in perceiving its qualities it would not follow that I erred as
to its position and relations ; nor if I erred as to its position need
I err as to its qualities. The complex of real and unreal elements
which makes up the object of consciousness may be analyzed
and tested part by part. Failure to attain certainty or to remove

300 A THEORY OF TRUTH AND ERROR

error in one respect will not make everything false or even doubt-
ful.

6. Summary. — Our effort in this third and final division of our
inquiry has been to explain the conditions under which truth and
error arise. Presupposing the conception of the meaning of truth
and error outlined in the first division, and proceeding on the basis
of the theory of consciousness set forth in the second division, we
have attempted to demonstrate in terms of the ' epistemological
triangle ' how the resultant system of cerebral currents and memory
traces may sometimes and in some respects imply real things and
relations outside itself, and how it may at other times and in other
respects imply the unreal contradictories of these.

As we said at the outset of our discussion, the problem of truth
and error has a peculiar bearing upon the controversy between
realism and subjectivism. Realism flourishes naturally as long
as man attains truth. It is the fact of error that has led to
subjectivism. Indeed we might say, without intending an im-
pertinence, that subjectivism is founded upon error and that
realism is founded upon truth. Error is as real as truth, but its
reality is not incompatible with the reality of the world.

THE PLACE OF ILLUSORY EXPERIENCE IN A
REALISTIC WORLD

I.
r

THE PLAC* OF ILLUSORY EXPERIENCE IN A
REALISTIC WORLD

BY EDWIN B. HOLT

ILLUSION, hallucination, and the erroneous experience in general,
we are told, can have no place in a universe where everything is
non-mental or real : and they cannot be satisfactorily accounted
for by realistic philosophy. Such is the challenge that has been
repeatedly thrown out from the idealist to the realist camp. And
although it has recently been taken up and admirably answered
in two all but impeccable articles by Professor Alexander and Mr.
Nunn,1 I propose to take up the issue once more and to add some-
thing if possible to the measure of satisfaction already vouchsafed
by these two gentlemen. A closer definition of the terms used in
the dispute is not necessary at the outset, for they are offered and
may be accepted as the current names for fairly unambiguous
phenomena. Sharper definitions, however, for these and other
terms will emerge in the course of the argument.

ILLUSIONS OF PERCEPTION AND THOUGHT

Erroneous experiences have been assumed to come under four
heads, according as the error is one of space, of time, of ('second-
ary') quality, or of judgment (thought).

1. Errors of Space. — An object is frequently seen as nearer
or farther, as larger or smaller than it really is; it may be seen in-
verted in position or distorted in shape, it may be seen double
or triple (spatially reduplicated) ; and the same frailties attach
to the senses of hearing and touch and to other modes of appre-

1 Alexander, S., On Sensations and Images. Proc. of the Aristotelian Soc., 1910,
N.S. 10, 1-35. Nunn, T. Percy. Are Secondary Qualities Independent of Per-
ception ? Ibid., 191-218.

303

304 ILLUSORY EXPERIENCE;

hension, such as that, for an instance, in which the semicircular
canals are a contributory factor. The person sees what is not there,
hence the act of seeing is constitutive in the case, hence the illu-
sory object (however it may be with a correctly perceived one)
is essentially mental and subjective. Thus if the two eyes are
sharply converged and then one eye is closed, near objects (far
ones are now out of focus) are seen as both nearer and smaller
than ordinarily, or than they really are. ('Really are' just
happens not to be a realist's phrase, as shall be explained later,
but I shall use it and still refrain from quotation marks out of
deference to the opponent, who, to judge by his actions, is unable
to state his case against realism except he be granted handy ac-
cess to some things that 'really are.') Now there is a machine
for manufacturing the lasts on which shoes are made. A model
last is placed in contact with one end of an arm and the machine
at once carves out of a block of wood a second last which is like
the model. The machine at work has quite the air of seeing its
model. Indeed, the comparison between duplicate and model
has an uncanny resemblance to the subject-object relation. How-
ever that may be, there is an adjustment which can be effected
such that the duplicate last is turned out smaller than the original,
but otherwise still like it, and the same adjustment brings any
point that shall be geometrically denned as the center of the last
nearer to the important part of the machine, the cutting edge.
I forbear to press the analogy : the essential point is, merely, that
a mechanical manipulation of the eyes which brings things nearer
and makes them smaller argues nothing for mentality or subjec-
tivity, for there is another machine at hand which can be as readily
manipulated with the very same effect. And in the shoe-last
machine I imagine that our opponent, so far from discovering the
subject-object relation and so forth, will even feign to evade the
comparison.

However, our next case is his own favorite out of a not too-
varied repertory. If a person presses one of his eye-balls out of
place, and keeps both eyes open, he will see double. He sees,

SPATIAL ILLUSIONS 305

for instance, a second mare's nest which isn't really there. This
demonstrates a remarkable creative function inherent in mental
process. A too literally-minded person might ask if it did not
somehow depend on there being two eyes, and this the more as
the direction given for setting this creative function of mind in
operation was distinctly physical, i.e. a jab on the eye-ball. In-
deed, it could almost be said to be directed on the eye rather than
addressed to the mind. This notion becomes even more plausi- 1
ble when one remembers that the stereoscopic camera habitually |
sees double. The notion becomes a certainty from the circum-
stance that in the case of a person with two eyes of which one is
blind, with, however, a creative mental apparatus equal to any 1
the best you may find, — in such a case the charm works not at
all. (It does not work even if a normal-sighted person closes one
eye-lid during the experiment.) This is proof by the Baconian
canon, so long as our opponent adduces only two factors — the
pair of eyes and the mental process. Or, to return to concrete
experience, I will ask the opponents of realism this : Do they in-
deed think that of the two images of one object which are given •
in the stereoscopic camera one of the images is 'illusory' be-
cause but one outer object is really there? Does the camera
lie ? If not, why is one of the two images on occasion given in
human vision, any the more illusory? "Ah, but," is the reply,
"no one asserts that the image in the camera is the object out in
front." To which I say that realism asserts just this, — the
image is genuinely (a part only, but a true part, of) the object
photographed. The professional photographer asserts it, saying,
in the manner of his trade, "We have caught your exact ex-
pression, Madame." And we all speak, as Robert Louis Steven-
son spoke, of "just that secret quality in a face that is apt to slip
out somehow under the cunningest painter's touch, and leave
the portrait dead for lack of it" ("An Autumn Effect"). The
physicist or astronomer absolutely asserts that his photograph
of a spectrum is integrally the spectrum under study. Now the
realist contends that when we say such things we mean them;

306 ILLUSORY EXPERIENCE

the opponent, that we do not. Yet everybody except this per-
verse school of philosophers continues to assert the same in any
case whatsoever of ' reproduction ' in art or craft ; and the philoso-
phers of that school assert it the instant they have doffed their
professional perruque and descended from the cathedra. Whether
they admit this or not, the point which they must admit is that
reduplication, if only of 'images,' is a common feature of purely
I physical systems, and that therefore the occurrence of redupli-
cation as a function of the human organs of sense argues nothing
for subjectivity lying behind that sense. The relation of 'image'
to the object and then to knowledge we shall presently examine.

Or again an astigmatic eye distorts its object ; so does a roughly
cast glass lens, and so do the innumerable facets on the surface
of troubled waters. A stimulation of the two sacs in the inner
ear translocates one part of perceived space within an otherwise
perceived space; just so every mirror and other reflecting sur-
face makes a new translocated space, which is in all ways geometri-
cally viable and is within the original real space : and so on for
every one of the spatial illusions and hallucinations.

Our opponent must withdraw from his citation of cases in
which the organs of sense yield multiplied or distorted images,
because of the invariable parallel to strictly physical systems.
And he must stand on a ground further back ; which would be,
as several of our opponents have indeed made their particular
point, that not the distorted image as such, but the distorted
image which asserts itself to be, or which the realist asserts to be the
real object, — that this is the crux for realism. In order to be
fair I hasten to add that not a few of the opponents of realism have
not to retreat to this ground, since it is their original position.
And I am only too eager to meet them there, as I shall do in due
course. Meanwhile many of our opponents have undertaken
to occupy the ground which I claim that we have just covered,
and all opponents are to be reckoned with. Indeed, it remains
true that in nearly every case where the idealist has ventured
within the somewhat unfamiliar and ticklish region of the concrete,

SPATIAL ILLUSIONS 307

he has brought against realism the charge which I have just par-
tially refuted. I say 'partially,' not because I account the physi-
cal parallels of illusion as inconclusive, but because we have so
far considered only the spatial cases. We now proceed to the
next group.

2. Errors of Time. — An object is not only frequently, but in-
variably seen at a moment of time later than that when it had the
position and other circumstances which it still has in our vision
of it. The illustration hallowed by the tenderest association for
the idealist seems to be the case of seeing the sun or other heavenly
bodies some millions of years behind time, or indeed millions of
years after it may have ceased really to exist. But now what
advantage over us has the photographic plate, as plainly physical
and as little mental or illusory as we all grant that to be ? And
if it be paradoxical and for realism ominous that we can see a
real thing long after it has ceased to be real, how much more
paradoxical and ominous must it be for physics if it is obliged to
concede that each smallest physical object remains under the
direct and real influence of other physical objects which have not
been existent or real for millions of years. It seems to me that this
argument proves either all or nothing. It may, I fancy, open
up magnificent vistas to theosophists and clairvoyants; but my
imagination halts when I try to conceive what it can open up to
a sane and sober philosopher. Nevertheless such a passage as
the following is characteristic of the greater number of writers
who are just now so superbly stooping to take notice of realism.
"The 'real' object always exists earlier in time than the perceived
object, which we hastily assume to be the 'real' object, yet which is
really but an element in our own experience, and not the ob-
ject, or eject, which exists in and for itself." l Succinctly stated,
the astonishing argument is that the mental image is not a part
of the real world, but is distinctly non-physical and non-real,
and belongs intrinsically to another, the subjective order of being,

1 Drake, D., The Inadequacy of 'Natural' Realism, J. of Phil., Psychol., etc.,
1911, 8, 371.

308 ILLUSORY EXPERIENCE

because it lags in time behind the real order of things : and this
is urged notwithstanding the fact that every case of image which
can be cited from the strictly physical world lags in exactly the
same way behind its real physical prototype.

Some of our opponents, as hi the previous case of spatial illu-
sions, stand on the safer ground of affirming that not the delayed
image as such, but the delayed image which asserts itself to be,
or which the realist asserts to be the real object, — is the crux
for realism. Thus one critic affirms that "In the face of inter-
vening machinery (of sense-perception) and the lapse of time,
realism cannot persist in saying that the very object thrusts it-
self in amongst the contents of consciousness." We shall consider
this latter argument in due course, but for those who find some-
thing in the former argument I wish to suggest an illustration
which they may find useful in their future animadversions on this
point, since it will be more novel than the somewhat threadbare
case of seeing the non-existing sun, and furthermore in their
eyes more damaging to the cause of realism. One discovers in
the psychological laboratory a kind of lapse of attention dur-
ing which sensory stimulations are prevented from reaching the
sensorium, or from coming to consciousness; yet if the lapse is
brief, say not over five-tenths of a second, sensory stimulations
given during the lapse reach consciousness afterwards and are
then perceived in the reverse of their true time order. This is, of
course, hard on realism, although the same reversal happens to a
bunch of delayed telegrams, letters, or express parcels.

3. Errors in Secondary Qualities, — The detachment of second-
ary qualities from physical objects has seemed so conspicuous
to the makers of history, that these qualities are rather generally
deemed to exist in minds and nowhere else, to be in very essence
mental. On the other hand, there is a group of thinkers who
assert that the secondary qualities are not in minds at all, neither
ideas nor even sensations being endowed with these qualities. In
this case either they are spoken of as if they were a part of the
physical objects, or else no status or habitat whatsoever, so far

TEMPORAL ILLUSIONS 309

as I have been able to discover, is assigned to them. I take
it that in urging the case of secondary qualities against realism,
our opponent assumes the case (for him) contrary to fact, that
the qualities are out there on the objects, and undertakes to argue
to a reductio ad absurdum.

The larger part of illusory instances here hi question needs but
brief comment, for their case is still parallel to the cases already
considered. As one passes down a dark street and past a lighted
window, the darkness of the street seems more profound, although
the night is really no darker : but just so a photographic plate
carried past the light of the window is thenceforward less sensi-
tive to the faint illumination of the darkness beyond, and this
illumination is for it one of reduced intensity. But in such a
simple fact of modified physical process no one sees raised at all
the issue between reality and unreality, or between the material
and the mental. The hill and wood that one looks out upon are
practically invariable in their chemical properties, yet as the day
progresses they are seen in a perpetual variety of light, shade, and
hue : but their luminous properties, although invariable, are invari-
able functions of the incident illumination, functions whose particular
values vary therewith ; so that the light that is reflected is hi reality
ever changing. Since it is reflected light that is perceived in this
case, and not the invariable functions, it would be a falsification
of reality if the light were not seehi in perpetual flux. The ortho-
chromatic moving film will record this diurnal flux in an entirely
parallel way. The change of 'appearance' as an object is carried
from peripheral to foveal vision is only a case of one light differ-
ently affecting differently sensitized surfaces: and the 'brightness
distribution' of the totally color-blind eye is photographed out-
right by exposing a tube filled with a solution of visual purple
from the eyes of frogs, to the several wave-lengths of the spectrum.

A feeble tone sounds fainter when a stronger tone is sounded :
so too it will yield less physical sound when occurring simul-
taneously with other sounds (owing to partial interferences). The
intense stillness after explosions has its counterpart in the ' fatigue '

310 ILLUSORY EXPERIENCE

of physical mechanisms. In binaural diplacusis one given pitch is
heard by the two ears as two different pitches (generally less than
one third of a scale-interval apart) : but so will one tone set up
overtones of different pitches in as many resonant bodies ; as one
may note by listening to several telephone transmitters which
are connected with one receiver. It may be alleged that audi-
tory 'beats' falsify the perception of physical sounds, but 'beats'
take place in every damped membrane, and are in human audi-
tion nothing but the true perception of what is going on in the
tympanum of the middle ear. Likewise, in some measure, at
least, auditory ' attention waves ' are the true record in conscious-
ness of interrupted function of the tympanum, owing to rhyth-
mical changes in the latter's tension, which prevent it from
vibrating responsively to some pitches. (Similarly visual attention
waves are in part at least due to like rhythmical changes in the
shape of the lens.)

The phenomena of dizziness and other peculiar affections in our
sense of orientation have found a singularly satisfactory explana-
tion in the Mach-Breuer theory, wholly in terms of the inertia
and elasticity of bodies. And a model has been constructed
which manifests in the real world many of the very same false
responses to changes of orientation that the human being per-
ceives by virtue of the labyrinth of his two ears. An almost iden-
tical organ has been found in plants. One can only regret, in
this connection, that none of our sense-organs contain gyroscopes,
for then there would be material startling enough to prove to the
satisfaction of some a dozen anti-realisms.

The case of all the other senses is exactly parallel, and I will
mention only the sadly overworked instance of the bowl of water
which feels hot to one hand and cold to the other, adding merely
that if our learned opponents who find this experiment so sig-
nificant will heat one thermometer and put a second for a time on
ice, the first will record as objectively as one pleases that the
bowl is really cold, while at the same time the second thermome-
ter will record that the bowl is really hot, In connection with

SECONDARY QUALITIES 311

all of the phenomena so far cited, the admirable papers of Profes-
sor Alexander and Mr. Nnnn, already referred to, will be found
most illuminating.1

But I have not yet mentioned hallucinatory secondary phe-
nomena, cases in which a color or sound comes in all distinct-
ness to consciousness, while nevertheless no such color or clue
to it is out there in the physical world. In a way positive, nega-
tive, and complementary after-images seem to belong here; and
yet they are really cases like the foregoing. A positive after-
image has of course a thousand parallels in the physical world,
and indeed it depends certainly on the nervous 'after-discharge/
which has been well made out by physiology. The image is
therefore a true perception of a real process, which in its turn
constitutes a part of the real properties of a real thing (as shall be
more amply brought out later). Likewise a negative after-image
is paralleled by the photographic 'negative.' The comple-
mentary after-image seems indeed a purely subjective hallucina-
tion. The yellow candle-flame is not blue as it looks in the after-
image. Vision after santonin poisoning shows objects, again,
suffused with an illusory or hallucinatory yellow cast. This latter
case is nearly paralleled by fluorescent bodies which increase the
wave-length of all light that comes to them. It is not probable
that santonin vision depends on fluorescence (although the visual
purple of the retina happens to be a fluorescent substance), but
it would be explained in physical terms, and given a real status
like that of our other instances, by any physiological process which
should have the effect of a yellow light-screen interposed between
the object and the sensorium. Although the means for this have
not been discovered, it too clearly rests on a subtraction from con-
sciousness (as it were a selective eye-lid) to be of interest in the
present connection.

The complementary after-image still looks more like a subjec-
tive creation. Now it will be recalled that Helmholtz plotted a

1 Cf., regarding these and similar cases, Dewey, J. of Phil., Psychol., etc., 1911,
8, 393-397.

312 ILLUSORY EXPERIENCE

curve showing the wave-lengths of the complementary color-
pairs,1 and I have learned from Professor G. W. Pierce that when
the capacity of the receiving mast of a wireless telegraph system
is tuned to a given length of Hertzian wave, it is ipso facto
tuned to a second wave-length as well ; and for each new tuning
it becomes sensitive to a similar pair of waves: that, further,
the curve of these wave-pairs shows precisely such a function as
the plot of the complementary color-pairs. The complementary
or antagonistic colors are comparable, then, not to the octaves
in audition (as some serious but abortive speculations have tried
to show), but to the complementary pairs of Hertzian waves.
Now it is not known how the cones of the retina pick up light-
waves and send them on as nervous impulses, but it is improbable
that they do this by virtue of any photo-chemical substance (as
is the case with the rods), for in spite of repeated investigation
of the cones no such substance has been found. (The question of
'visual substance/ as raised by Helmholtz and Hering, is not
necessarily involved here.) But it is most probable for a variety
of reasons which have been advanced by Meisling,2 that the cones
resonate to waves of light, as in Helmholtz's theory of audition
the radial fibers and organs of Corti are believed to resonate to
sound. Now light-waves and Hertzian waves are closely related
physically, so that if the view of Meisling is correct, — and it is
not only the most reasonable view so far offered, but it is itself
well supported by facts, — the complementary colors of vision are
nothing but a true presentation in consciousness of the fact of
complementary attunement of light-resonators. This is not an
assured fact, but it is a view possessing too high a degree of
probability to leave any interest attaching to the case of com-
plementary colors as a special difficulty in the path of realism.
The case of pure hallucination involving no peripheral stimu-

1 Helmholtz, H. von, Handbuch der Physiologischen Optik, Hamburg, 2te
Auflage, 1896, 317.

1 Meisling, A. A., Uebcr die chcmisch-physikalischen Grundlagen des Sehena.
Zsch. /. Sinnesphysiol., 1907, 42, 229-249.

SECONDARY QUALITIES 313

lation whatsoever, the case, that is, of centrally Induced sensations
and images, or, as Johannes Miiller called them, "fantastical
visual phenomena," is more unmistakably to our purpose. And
I confess that this is the first of the arguments allotted to me in
which I can see enough plausibility to enable me to meet the adver-
sary with either amiability or patience. I heartily grant the pro-
priety of our opponent's question, How can realism pretend
to assert the reality of the color, sound, and perhaps tactile or
olfactory sensations which are vividly present in the dreams of a
person sleeping, it may be, in a box no bigger than his coffin ? The
case has still two aspects : first, how can these purely hallucinatory
secondary qualities have, even in themselves alone, any sort of
being other than a subjective and mental being? Second, the
argument already twice referred to, how can they pretend to
assert themselves to be, or how can the realist pretend to assert
them to be the real object ?

The former difficulty first. It must have occurred even to the
idealist, in a ruminative moment, what a singular thing that
cleavage in our universe, the sharp division between primary and
secondary qualities that is ordinarily attributed to Locke, after
all is. Whether one views our 'universe' as a material system,
a scheme of relations, a subjective realm, an 'Absolute' sys-
tem, or whatsoever, one is ever confronted by the strange apart-
ness of the secondary from the primary qualities, a cleft that is
bridged only by the bare one-to-one correlation of some of the
colors, sounds, and so forth with certain vibratory wave magni-
tudes found in the primary system. Since this is a bare one-
to-one relation, not even for instance admitted to be a true causal
relation, the case presents even to the thought of our own day
as clean and gratuitous a 'preestablished harmony' as was ever
dreamt of by Leibnitz. This cleavage and the mystery of error
constitute, I take it, the chief empirical motives (as opposed,
say, to temperamental motives) which have led thinking men
into some form of subjectivism. Since, however, subjectivism
by no means clears up this inscrutable correlation, it remains to

314 ILLUSORY EXPERIENCE

me personally a mystery that, so far as I know, no explanation
has been attempted of this harmonious accident — so uniform
and inveterate. But is this fixed correlation a fixed accident?
This is a question which must be answered on solely empirical
grounds, if our adversary grants that term to have any meaning,
and without regard to realistic or anti-realistic prepossessions.
For the empirical psychologist the unique and isolated position
of secondary qualities is enunciated in Johannes Miiller's doc-
trine of the specific energy of nerves, for which Miiller himself
gave various formulations, one of which is substantially the theory
generally believed to-day. The germ of the doctrine descends
to us from Locke through Hume to Kant, for it was Kant's epis-
temology that Miiller, by his theory, aimed to exhibit in concrete
operation. (He hoped thus to confirm Kantian metaphysic, and
although the work turned out, unconsciously to Miiller, a carica-
ture, we still note among idealists a real affection for the 'Son-
derstellung der sekundaren Qualitaten.') Nevertheless, explicitly
Miiller proceeded empirically, and based his theory on two groups
of phenomena; the 'fantastical visual phenomena,' or in general
the central excitation of mental processes; and the fact that
some nerves when stimulated by some other than their normal
stimulus still respond by yielding the same sensation as when
exposed to their normal or 'adequate' stimulus. Thus, it was
believed that a surgical severing of the optic nerve produced a
confusedly brilliant burst of light and color sensations. Clearly
the world outside the body contained only primary qualities and,
more pertinently, vibration-rates, while the secondary qualities
were certainly found only in the nervous system and undoubtedly
(but this was pure assumption in the supposed Kantian interest)
only in the mind. And it was a specific energy of nerve fiber
'or' brain center (the latter being the modern view), that, al-
though having nothing about it that resembled a secondary
quality, had the power of arousing the secondary qualities in the
mind. Thus the specific energies were the physiological counter-
part of a Kantian category.

SPECIFIC ENERGIES 315

But obviously the specific energy of the optic nerve must differ
from that of the acoustic nerve in order to account for the differ-
ent effects of the two on the mind. Miiller spoke somewhat
roughly of specific energies, qualitatively1 differing, for each of
the several senses. The assumption was later made by Helm-
holtz, and not without reason, of a specifically different energy
not merely for the modes of light, sound, heat, and so forth, but
for many of the qualitative differences within these modes as well.
Thus we have the sundering of primary and secondary qualities
with the harmonious accident of their correlation, restated in the
terms of mind and body or rather of mind and the cortical cells
of the cerebrum, with nothing done to explain the steady correla-
tional accident. (And as for the Kantians, it is hard to see how
they can feel grateful to Johannes Miiller for exhibiting so defi-
nitely the slender cortical threads by which the mind is held in
touch with a contrastingly non-mental world, even though the
latter was devoid of secondary qualities.)

The point about this theory which interests us now is that nerve
physiology has not been able to discover any trace of specific
nerve-energies.2 All nervous impulses seem to be of quite the
same ' qualitative ' order, differing only in intensity : sensory im-
pulses are even not different from motor impulses. Yet the
question remains, — How can the mind (to continue Miiller's con-
ception of the case) be affected to different modes of sensation by
nervous impulses which are all qualitatively alike? And it is a
question not merely of understanding the cases of centrally in-
duced hallucinations and of normal sensations caused by abnormal,
'inadequate,' stimulation, but a question of still broader import,
— How can any physical stimuli whatsoever, whether normal

1 Miiller uses the term here with no reference to ' secondary qualities,' but
merely as one might loosely speak of a ' qualitative chemical process.'

* The unsuccessful attempts in this direction, notably by Hering, have chiefly
been actuated by the theoretical needs of sense-psychology. Cf. Boruttau, H.,
Alte und neue Vorstellungen iiber das Wesen der Nervenleitung. Zsch. f. aLlg.
Phyaiol. (Verworn's), 1902, 1, Referatenabteilung, 1 (especially 10-11). Hering,
E., Zur Theorie der Nerventhatigkeit, Leipzig, 1899 (especially 8, 10).

316 ILLUSORY EXPERIENCE

or abnormal, produce different qualitative sensations when the
differences between the stimuli, the differences of wave-length
or vibration-rate, are merged and lost on their way to the brain
in nervous impulses which appear to be all alike.

The readiest answer should seem to be that the differences re-
quired are not differences in impulse but in nerve-fibril : each nerve
fibril different from every other might, whenever excited, give the
cue to the mind for one specific secondary quality. But no such
differences between nerve fibrils have ever been discovered ; any
more than between nerve impulses.1 Might then the differences
be in the cortical cells at which the fibrils terminate? This has
been somewhat investigated and a very slight difference in the
chemical composition of different lobes of the brain found. Even
this has not been confirmed, and I believe that no investigator
has ever claimed to find chemical, histological, or other differen-
tiations between homologous cells of one lobe ; as our assumption
requires. (Differences between the cells of different cortical
layers will not serve our purposes, since these differences between
layers extend practically unmodified over the whole cortex.) Pro-
fessors Sherrington 2 and McDougall 3 have shown that not the
cell-bodies but the synapses between neurones are in other respects
(such as duration and fatigue phenomena) significant for conscious-
ness; but here again no differences between synapses even re-
motely approximating what we require have been adduced. The
only course remaining is that taken by Professor Mtinsterberg
in his 'action theory,' 4 where the correlation between neural im-
pulse and secondary quality in the mind is conceived as a bare

1 This distinction is at best a somewhat academic one, since it is hardly conceiv-
able that any significant differences in nerve-fibrils should not bring in their train
phenomena that would become manifest in the experimental study of the nerve-
impulse.

1 Sherrington, C. S., The Integrative Action of the Nervous System. New York,
1906, 14 ff. Cf., however, earlier writers such as Goldscheider, A., Die Bedeutung
der Reize im Lichte der Neuronlehre. Leipzig, 1898, I.

1 McDougall, W., Physiological Psychology, London, 1905, 30-33.

4 Milnsterberg, H., GrundzUge der Psychologic, Leipzig, 1900, 1, 15.

SPECIFIC ENERGIES 317

one-to-one correlation, no other differentiation in the nervous sub-
stratum being suggested than one of spatial position. And this
view again brings us back to the unexplained preestablished har-
mony ; and to the mind of a realist, at least, nature abhors a pre-
established harmony. But another and strictly empirical reason
makes impossible the assumption which we have just considered.
In the words of the late Professor Nagel,1 "Unquestionably an hy-
pothesis which should permit us to ascribe to individual nerve-
fibers a single qualitatively determined and unvarying mode
of excitation (Erregungsart, and the context shows that he is
referring not to the form of stimulation, but to the nature of the
sensational response) would be by far more satisfactory, and would
stand in closer accord with other investigations. The special
investigations in sense physiology, however, are not at present
favorable to such an assumption." The evidence is as follows.
Miiller left his theory in a crude form, inasmuch as he thought of
a 'specific energy' for each 'sense' — sight, hearing, and so
forth; but each 'sense' gives us several qualities (as four for
taste, at least several thousand for hearing, and an indefinitely
large number each for sight and smell) ; and every argument that
speaks for specific energy at all for the several sensory modes,
speaks just as imperatively for a separate energy for each quality
within the mode. This ' undoubtedly influenced Helmholtz, in
his theory of audition, to assign to each fibril of (the cochlear
branch of) the eighth nerve, the work of transmitting a separate
auditory sensation of pitch. The fact that these fibrils are con-
nected serially with the graded series of organs of Corti made the

1 Nagel, W. A., Die Lehre von den spezifischen Sinnesenergien. Handbuch der
Physiologic des Menschen, Braunschweig, 1904, 3, 15. Professor Nagel was not
optimistic, indeed, about several of the reputed facts on which Miiller based his
theory. "So far as I know we have still no proof that mechanical or electrical
stimulations of the optic nerve-trunk produce sensations of light" (7). And
again, " The most clean-cut (eleganteste) and in fact the only really clear confirma-
tion of Muller's law (regarding the results in sensation of ' inadequate ' stimula-
tion of nerve-fibers) is the already mentioned experiment on the chorda tympani
of the exposed middle ear ; mechanical, chemical, and electrical stimulation of the
central end of the severed nerve does produce a sensation of taste " (8).

318 ILLUSORY EXPERIENCE

theory very plausible. Helmholtz was similarly influenced (al-
though the evidence is here less obvious) in his assumption of
three fundamental qualities of vision. Now the extension of the
theory of specific energies to the sensory qualities, as required if
the theory is to explain the facts which it undertakes to explain,
has been thoroughly unsuccessful, and has been put by recent
investigations quite out of question.1

In the first place the theory never has, in spite of several efforts,
been extended with any degree of completeness to the qualities.
The greatest success was attained in Helmholtz's theory of audi-
tion. This theory conceives the series of pitches as the quali-
tative auditory series (of which the justice has been disputed),
shows the different pitches to be carried to the brain by different
nerve-fibrils : and in spite of some doubts on the score of differ-
ence- and summation-tones was on the whole a credible theory.
The theory still meets with several difficulties, nevertheless, a
rather serious one being the difficulty of conceiving how the radial
fibers with their almost microscopic dimensions can 'resonate'
to so great a range of pitches as are included in the audible range
of hearing, and specially how such resonance is possible between
the little fibers and the lower audible wave-lengths which are
little short of sixty-four feet. Furthermore these radial fibres
are transversely bound together, so that any true resonance must
be made more difficult, and a specific resonance of individual
fibers to respective individual tones seems quite out of question.
Were this not the case, it is estimated that the radial fibers would
be numerous enough to correspond to the number of distinct au-
dible pitches if this latter is taken as the same as the number of
successive pitch differences discriminable in an ascending or de-
scending continuous series of tones. In fact, however, the se-
ries of pitch sensations is continuous, and so includes infinitely
more pitches than are brought out in the step-wise series yielded

1 Nagel, W. A., op. cit., 14. Professor Nagel's entire essay is of interest in thia
connection. The history of Miiller's theory is ably treated by Weinmann, R., Die
Lehre von den spczinschen Enorgien, Hamburg, 1895.

SPECIFIC ENERGIES 319

by the discrimination experiment. Helmholtz himself admitted, if
I recall aright, that many intermediate pitches must be accounted
for by the simultaneous activity of adjacent organs of Corti
excited severally in differing degrees of intensity. This rather
stretches a point in order to prop up the general theory of a nerve-
fibril to a sensory quality : and yet one is tempted to let it pass,
owing to the close resemblance between near-lying pitches.

But in vision the case is less favorable. The color series is also a
continuous one, and since on the red-green zone of the normal
retina (a very considerable area) every hue can be perceived at
every point, it was from the outset out of question to look for,
or to assume, a cable of nerve-fibrils going to every point and large
enough to provide a separate fibril for each of the discriminable
color differences, to say nothing of the actually continuous
series of such differences. So here Helmholtz stretched still fur-
ther (and very far) the point which just now we let pass. He as-
sumed, in effect, three fibrils going to what is practically each
point of the red-green zone, and assumed one of the three colors
red, green, and blue to be aroused in the mind by activity of one
or other of the three fibrils. Intermediate colors, as in the du-
bious point conceded in the case of audition, were aroused in the
mind by the simultaneous excitation in different relative amounts
of two or even three of the fibrils. Thus yellow, in which few
persons recognize any resemblance to either red or green, must
nevertheless be understood as the mental result of simultaneous
and equal excitations of red and green fibrils : and white, which
resembles no color, as the mental result of equal excitations of
red, green, and blue fibrils together. Now here surely the matter
has been stretched too far, for if yellow and white are not sufficiently
distinct qualities to have special fibrils assigned them, then the
theory of a nerve-fibril to a sensory quality has simply failed.
The case is furthermore worse than in audition, for the mind must
be inexplicably aroused to perceive different qualities on the ex-
citation of different fibrils, and intermediate qualities according
to relative and exactly discriminated quantities of excitation;

320 ILLUSORY EXPERIENCE

and now, per contra, it must be aroused to apprehend the same
quality on the excitation of several thousand different red, or
green, or blue fibrils. When the point is stretched so far, the
principle of a nerve-fibril to a quality is squeezed to vanishing;
it is lost.

Moreover, continuous qualitative series quite bridge in many
cases the gaps which the older psychology rather jauntily assumed
to exist between the 'five senses.' There are innumerable sen-
sations which are aroused by the simultaneous excitation of
taste and smell fibrils ; indeed, all but four of the so-called tastes
are due to the cooperation of olfactory excitations, and yet it
is notorious that none but the trained, and few of these, can men-
tally resolve these sensations into taste plus smell. For nearly
every one they are as unanalyzable as is white unanalyzable into
red, green, and blue. Again the variety of tastes is greatly in-
creased by the (mentally) unsuspected assistance of the touch
sense. The senses of touch, pain, warmth, and cold (known to
have separate end-organs) are inextricably co-involved in the
production of a vast variety of 'dermal' sensations; while, ac-
cording to Professor von Frey, the sensation of heat is due to the
simultaneous action of organs which if excited severally would
yield sensations of warmth and cold. The number of instances
could be increased to any length, but I will further mention but
one sensation which I may call the still, small voice of psychol-
ogy : the sensation which has no namable quality. It is familiar,
I think, to every trained introspector, and certainly to every ob-
servant one; one seeks in vain to assign it to any of the famil-
iar senses, and it is distinctly not visceral or otherwise of a
proprio-ceptive order. Thus I have several times been as assured
as the nature of the case allowed, that the stimulation which
yielded the experience was either visual or auditory, and still I
was quite unable to recognize the result in consciousness as either
a color or a sound ; yet the sensation was moderately intense and
it held its own well under the process of introspection. By way of
introspective description I can only say that its quality suggests

QUALITATIVE SERIES 321

something primal and unelaborated. I mention this at some
length because it will interest us later : and also because it is so
little suggestive of a rough and ready scheme of specific energies.

And thus it is that the theory of specific energies of nerve,
twist and refine it as one will, encounters so many difficulties and
ends up with such a bland profession of ignorance, that one must
go back, and indeed for economy of thought much prefers to go
back, to the original facts, perplexing as they are. The considera-
tions which I shall now adduce seem to me to plunge the merci-
ful bodkin into the Miillerian theory and also to throw a flood of
light on the seemingly baffling facts.

If nerve physiologists have been unable, for the support of
Miiller's theory, to discover anything like a qualitative difference
among nervous impulses, nor yet suitable chemical or histologi-
cal differences among cortical cells and synapses, certain recent
discoveries have shown something about the nervous impulse
which both does away with old conceptions and introduces rather
extraordinary new ones. That something is, that the nervous
impulse, and particularly the sensory nervous impulse, presents
periodic fluctuations of a frequency vastly higher than was hitherto
suspected. In the words of Professor Sherrington,1 "The number
of separable excitatory states (impulses) engendered in a nerve-
trunk by serially repeated stimuli corresponds closely with the
stimuli in number and rhythm. Whether the stimuli follow each
other once per second or five hundred times per second, the ner-
vous responses follow the rhythm of stimulation. Using contrac-
tion of skeletal muscle as index of the response the correspondence
at rhythms above thirty per second becomes difficult to trace,
because the mechanical effects tend at rates beyond that to fuse
indistinguishably. The electrical responses of the muscle can
with ease be observed isolatedly up to faster rates : their rhythm
is found to agree with that of stimulation: thus at eighty per
second their responses are eighty per second. If the muscle note
be accepted as an indication of the respbnse of the muscle, its pitch

1 Op. tit., 42-43.

T

322 ILLUSORY EXPERIENCE

follows pari passu the rate of stimulation of the nerve through a
still greater range." "The refractory period in nerve-trunk con-
duction seems to last not longer than lo-." The last sentence is
equivalent to the statement that the periodicity of the nerve im-
pulse can go as high as one thousand per second; and it leaves
the question whether the frequency does not go higher, unpreju-
diced, for further investigation. Almost all of the recent work on
axis-cylinder conduction, muscle-tone, tetanus, 'Treppe,' and re-
fractory phase gives indication of a similar oscillation of the
nervous impulse at rates quite unsuspected by physiologists of
an earlier day.1 Thus, for instance, Professor Piper finds that in
voluntary contraction (under virtually normal conditions, then)
the flexor muscle of the lower arm receives fifty nervous impulses
per second.2 The reason why this discovery has been reserved
until so recent a date is of the simplest, and is the one hinted at
by Dr. Sherrington in the passage just quoted : the mechanical
inertia of the recording instruments hitherto used in studying the
nervous impulse has been such as to 'fuse indistinguishably '
anything but slow fluctuations of the nerve-current intensity.
But the mass and variety of evidence already on hand, together
with its unquestioned authority, make it certain that as the use
of finer electrical recording instruments progresses (particularly the
oscillograph, which is a relatively new instrument in physiological
laboratories), we shall become acquainted with a large field of
phenomena relating to high frequencies of the nervous impulse.

Now, as has so often happened, evidence of the same fundamental
fact has been simultaneously accumulating from another source,
that of specifically sense physiology or psychology. In 1907 Lord
Rayleigh 3 published a paper in which he showed conclusively

1 1 confess to being so far baffled in my search for the earliest experiments in
this line. It would be interesting to be able to assign the historical priority in what
is after all a profound modification of the earlier conception of the " nervous wave "
of, say, a second's duration or over. Cf. Bering, E., op. cit., 29.

* Piper, H., Arch. f. d. ges. Physiol. (Pflueger's), 1907, 119, 301; also ibid..
1909, 129, 180.

» Rayleigh, Phil. Mag., 1907, Sixth Series, 13, 214.

NEURAL PERIODICITIES 323

that sounds of lower pitch than 128 d.v. per second, at least,
(and perhaps of pitch as high as 512 d.v.) could not be localized
by means of differences between their heard intensities ; but that
they could be and were localized on the basis of their relative phases
as they entered the two ears. Lord Rayleigh adds, "It seems no
longer possible to hold that the vibratory character of sound
terminates at the outer ends of the nerves along which the
communication with the brain is established. On the contrary,
the processes in the nerve must themselves be vibratory, not
of course in the gross mechanical sense, but with preservation
of the period and retaining the characteristic of phase — a view
advocated by Rutherford, in opposition to Helmholtz, as long ago
as 1886" (loc. tit., pp. 224-225). Now earlier, but unfortunately
either not published or not adequately emphasized and there-
fore unnoticed, work had already been done which proved the
importance of phase-differences in the localization of sound, by
Sylvanus Thompson,1 Professor L. T. More,2 and Mr.. M. Green-
wood ; 3 and the fact has received further confirmation from sub-
sequent experiments of Lord Rayleigh,4 Professors Myers and
Wilson,5 Mr. Bowlker,6 and Professor More.7 The last named
author finds that phase-differences are effective up to pitches in
the vicinity of 1024 d.v. (loc. tit., p. 314). Professors Myers and
Wilson have tried to subsume the new facts under the old inten-
sity-ratio theory, by alleging that phase is operative only as it
is correlated with intensity, but their explanation is too patently
a case of ingenuity expended to save a preconception. If the
nerve-impulse were known not to show phases of the periods
here involved, one would perhaps let the explanation in terms of

1 Referred to, but place of publication not given, by Mr. Greenwood. (Pro-
fessor Silvanus P. Thompson intended ?)

1 More, L. T., Phil. Mag., 1907, Sixth Series, 13, 452.

1 Greenwood, M., Physiology of the Special Senses, London, 1910, 83.

4 Rayleigh, Phil. Mag., 1907, Sixth Series, 13, 316; and 1907, S. S., 14, 596.

1 Myers, C. S., and Wilson, H. A., Brit. J. of Psychol., 1908, 2, 363.

• Bowlker, T. J., Phil. Mag., 1908, Sixth Series, 15, 318.

» More, L. T., ibid., 1909, S. S. 18, 308.

324 ILLUSORY EXPERIENCE

the intensity-ratio theory stand, however improbable. Since,
however, the very periodicities which are prerequisite have been
independently proved to exist, the view of Lord Rayleigh and
Professor More is altogether the natural one. Professor More
pertinently says (loc. ciL, p. 319), "The only objection to the
idea that the ear is capable of detecting the phase of a sound
or at least the difference in the phase of two sounds, is that it is
difficult to reconcile with our theories of audition."

But the experiment showing that phase-differences govern the
localization of the lower pitches is one in which no 'hypothesis'
and indeed very little deduction has entered, whereas the Helm-
holtz theory of audition (to which Professor More must mainly
if not exclusively have referred) involves a grave hypothesis,
in the much-disputed resonance assumption, and was further-
more framed not a little in the interests of Miiller's 'specific
energies,' l — a theory which we have already seen to be value-
less. I say that it is the Helmholtz theory to which Professor
More must have referred, because of the three other prominent
theories, Ewald's in no sense depends on the Miillerian tradition,
and Rutherford's and Meyer's are irreconcilable with that tradition.
And the new facts regarding auditory phase are not difficult to rec-
oncile'with these theories. Indeed, Dr. Rutherford's theory consists
in little else than an interesting, and to my mind valid, protest
against specific energies, and in favor of the view that sensory
qualities are conveyed to the sensorium by vibratory nerve-im-
pulses whose rates are closely related to those of the impinging
physical stimuli. The considerations adduced by either Dr. Ruth-
erford or Professor Ewald are hardly complete enough to be called
a 'theory' of audition; those of Professor Max Meyer better de-
serve the title. This investigator has offered a theory which takes
into account and plausibly explains all the important peculiarities
of audition (specially some of those which Helmholtz found
most difficult), except the phenomenon of pitch ('quality'),
which generally is the first point that a psychological theory

1 Cf. Ebbinghaus, H., Grundziige der Psychologic, Leipzig, 1905, 335.

NEURAL PERIODICITIES 325

undertakes to explain. I cannot discover that Professor Meyer
has once made explicit mention of pitch, or how he understands
it to be transmitted to the brain. And yet his theory is posi-
tive on this point ; what it requires is to suppose that pitch is
transmitted not by specifically different nerve-fibrils (as in
the Muller-Helmholtz conception), but by nerve-impulses, along
any or all fibrils, which consist of periodic vibrations identical
in rate with the vibrations of the outer sound stimulus. I
have never understood what I must call Professor Meyer's mys-
terious reticence on this point ; unless, indeed, one may suppose
that, unaware of the arguments in his favor adduced by Dr.
Rutherford and many other physiologists, he has hesitated to
give explicit prominence to a feature of his theory which so widely
departs from the Mtillerian tradition. Indeed, the views of Meyer
and Rutherford are not rival, but complementary and entirely har-
monious theories. Together they form a compact, complete, and
very promising theory of audition ; which, for that matter, would
seem in no wise to jar with the fragmentary and somewhat whim-
sical speculations of Professor Ewald, although I can assign no
immense value to these latter. In short, I have no hesitation in
affirming that what we may call the Rutherford-Meyer theory is
an adequate theory of audition, and that in view of recent dis-
coveries in nerve physiology, it has the distinct advantage over
the theory of Helmholtz. I could substantiate this conclusion
in much greater detail, but in our present connection I have
wished only to show that all considerations require us to aban-
don 'specific energies/ l that the facts of nerve physiology point
unmistakably to the view that the quality of sensations is trans-
mitted to the brain by vibratory nerve-impulses, that certain
facts of sense psychology prove this in the field of audition, and
further that here the vibration-rate corresponds to that of the

1 It is interesting that in 1899 Hering in his defense (loc. cit., 8) of " qualitative
specific energies," cites even Helmholtz as an opponent of them. In other words,
Helmholtz believed that the specific energies were to be interpreted in some quanti-
tative fashion, possibly in terms of nervous vibrations (ibid., 29).

326 ILLUSORY EXPERIENCE

outer sound stimulus, and lastly that this is not "difficult to recon-
cile with our theories of audition."

How may such a view, now, fare in the field of vision ? — the
only other sense which has been at all exhaustively investigated.
The reply is that it fares very well. In a paper already re-
ferred to Mr. Meisling has presented very urgently the following
considerations regarding color (cone) vision. Heat, light, and
Hertzian waves belong to the same physical order, all being elec-
tro-magnetic vibrations and differing in the one essential of wave-
length or rate. These waves can and do affect substances photo-
chemically, as they, have hitherto been conceived to affect the
rods and cones of the retina; that is, by means of so-called visual
substances. But there must be a photo-chemical substance to
be affected. In the case of the rods such a substance is found in
the visual purple, and the conscious result is white or gray sensa-
tion. In the cones no photo-chemical substance has been demon-
strated, although the purple of the rods was discovered relatively
early (by Boll in 1876), and three or more visual substances have
eagerly been looked for in the cones. Now the 'visual substances'
were always purely hypothetical in our color theories, and were
conceived merely by analogy with the pigmented cells of rudi-
mentary vertebrate eyes and with the visual purples of the rods
in vertebrates. This analogy was not unreasonable, but it is
unreasonable to cling to it when almost a half century of investi-
gation has failed to bear it out, by demonstrating not even one of
the supposedly requisite photo-chemical substances. Now the
cones need not be conceived of as actinometers ; they may be
thermometers (bolometers), or resonators. They are probably
not thermometers, since then they would be more sensitive to
red than to yellow light, and still more to the slower heat waves
than to red. But the cones may very well be resonators. If they
are, the fact would account for their remarkable unfatiguability,
a point which Hering emphasized in his arguments against Helm-
holtz, and to provide for which (in part) Hering conceived his
antagonistic ana- and katabolic visual processes: a conception

NEURAL PERIODICITIES 327

which was applied by Hering to the temperature sense, and which
was overturned by the discovery of separate warmth and cold spots.
Now Meisling adduces facts concerning the structure of the cones,
their several types, the structure of the several layers of the retina,
the contraction of cones under light stimulation, and concerning
the rest-currents and action-currents of the retina, showing how
these facts well agree with the view that the cones are electro-
magnetic resonators.

Perhaps the most interesting confirmation of this view lies in
the complementary attunement of those instruments which re-
ceive Hertzian waves ; r a point which I have already touched on.
This would account inevitably for the complementariness of colors,
and in this regard both of the present theories of color are un-
satisfactory. Helmholtz's 'errors of judgment' are in many
phenomena of antagonism altogether far-fetched : and Bering's
kata- and anabolic processes are unconfirmed hypotheses; they
meet special difficulties in regard to the white-black series, and
they contravene sound physiological analogy. For, as has been
pointed out before, it is as unprecedented that an ana-bolic
process should directly yield sensation, as that it should directly
cause muscular contraction.

It is admitted on all sides that neither the Helmholtz nor the
Hering theory of vision had ever the authority of Helmholtz's
theory of audition, and that neither is to-day an acceptable theory.
It is generally believed that each contains errors so radical that
it would take elimination as well as supplementation to trans-
form it into a definitive theory. Such a transformation, of either
theory, Meisling offers when he suggests in place of 'visual-
substances' three or more differently attuned types of retinal
resonators. And, as we have seen, the same suggestion contains
the foundation for an explanation for the so-far unexplained an-
tagonism of colors. Whatever the result may be for color the-

1 Pierce, G. W., Experiments on Resonance in Wireless Telegraph Circuits.
Parts II, III, V., Physical Rev., 1905, 20, 220 ; 21, 367 ; 1907, 24, 152. Also Theory
of Coupled Circuits. Proc. Amer, Acad. of Arts and Sci., 1911, 46, 293.

328 ILLUSORY EXPERIENCE

ory in general, our present point is that we have cogent arguments
for believing that the visual impulse traveling along the optic
nerve is, as in the case of audition,1 a vibratory impulse whose
period corresponds with the vibration rate of the impinging
stimulus. And the opposing view of photo-chemical substances
with 'qualitative specific energies' has no direct evidence to
support it and a considerable body of evidence to refute it. Mr.
Meisling's position, I believe, is conservative, and more secure than
any other.

When we come to the other senses much less can be said on
either side of our question. I cannot discover that anything
pertinent, and I might almost say anything at all, is known of the
actual physiology of the remaining organs of sense. Certainly
all details of the processes of stimulation here are veiled in mys-
tery. These other senses are generally thought to be either me-
chanical or chemical as to their immediate stimulation. Rod-
vision remains a chemical sense of which no further details are
known concerning the nervous impulse that the bleaching of the
visual purple initiates. Yet the minute hairs on the olfactory
and gustatory cells have given rise to some speculation as to
whether the ultimate stimulus here might not be mechanical after
all. The temperature organs may well be resonators to heat
waves, but that is pure hypothesis. The pain, touch, and joint
senses are almost certainly mechanical in their mode of stimula-
tion, but that statement has no special significance for one who
is trying to understand the physiological processes involved in
their stimulation. The fact that mechanical stimulation of the
chorda tympani, a gustatory nerve which passes through the
middle ear, produces a sensation of taste, remains the one undis-
puted case in the series that was adduced by Miiller. It does not

1 Lest the reader think me perverse in arguing against resonance in audition
and for it in vision, I must remind him that I am arguing not in the interests of theo-
retical grace or symmetry, but solely on the basis of evidence. Audition might
very well depend on resonance and still the nervous impulse might be vibratory,
while still the Muller-Helmholtz hypothesis of a nerve-fibril for each auditory
pitch might be untrue.

NEURAL PERIODICITIES 329

prove the existence of a qualitative specific energy, but it may
squint that way, and must be borne in mind in a dispassionate
survey of the whole situation.1 We are woefully ignorant of all
these matters, and in science imagination is no counter-irritant for
ignorance. The Mtillerian theory of specific energies receives
no support from the total ignorance prevailing in this field, while
it is refuted by much that we do know about vision and audition.
If conjectures are to be made as to the nervous impulse in the
mechanical and chemical senses, the safest way will be to draw
some analogy with what we do know about the light and sound
senses. I personally believe that gratuitous conjectures (as
opposed to grounded deductions) are mischievous, and I have no
interest in them, whether they are pro- or anti-Miillerian. There
is a general consideration, however, which may properly be men-
tioned here. The specific energies, whether in nerve-fiber, brain-
cell, or synapse, are conceived of as differing qualitatively from
one another. Now science has so frequently found that examina-
tion of what appear to be qualitative differences shows them to
be really quantitative differences, that it has come to be a maxim
of science that there are no 'qualitative* differences. This, I
think, oversteps the mark : there are qualitative differences,
obviously, but these are usually analyzable into quantitative dif-
ferences ; precisely as every starch is a starch, but can be analyzed
into one of a series of quantitatively related carbohydrates. The
maxim should then be, simply, that quality is not an ultimate
category of natural science,2 and this maxim is universally ac-
knowledged save by a small group of persons of a certain, possibly

1 1 could mention several explanations which have been offered for this phenom-
enon, which expressly avoid the assumption of a specific energy, and which are
consonant with a vibratory theory of the nerve-impulse. I do not mention them
because the facts still seem too meager to point clearly to any explanation whatso-
ever, either Muller's or another's.

1 1 am speaking here of the more concrete empirical sciences, and do not by this
wish to prejudice the question as to whether philosophically ' quality ' is or is not
an ultimate category. The maxim referred to is the equivalent of — " Look with
suspicion on all apparently qualitative differences.'*

330 ILLUSORY EXPERIENCE

rather vague, type of mind, such as, for instance, the neo-vitalists.
Now it follows from this that the specific energies can hardly hope
(or desire) to retain their peculiar, strictly 'qualitative' status,
and I should like to ask the believers in specific energies if they
can think of any other quantitative interpretation for these
qualitative differences, than one which I have been arguing for.
If I have spoken against specific energies in toto, it was solely
because most of the followers of Mtiller cling tenaciously to their
unanalyzable 'qualities' and account them the palladium of
their sect. It is significant indeed that Helmholtz, who did more
than any one else to substantiate Miiller's theory by physiological
evidence, held that the energies differed only in quantitative ways.1
And it is only fair to add that in Miiller's own writings the word
'quality' is used only in a casual fashion, as if it slipped in while
his attention was focussed on other matters. I do not think that
Miiller anywhere proposes to exclude the resolution of these
'qualitative' into quantitative differences. And it is a matter
of indifference whether the arguments which I have presented
shall be said to oppose or to interpret and extend the original
theory of specific energies. The real purpose of this all too te-
dious digression into physiology, is to show that there is excellent
and abundant evidence of minute periodic fluctuations in the nerve-
impulse, and of a close correspondence, in the cases of audition and
color-vision, between these fluctuations and the vibration rates
of the impinging physical stimuli.

We have next to examine the secondary qualities from the
so-called subjective or introspective point of view: and here we
shall presently see the bearing of the foregoing discursus on the
realistic issue.

Our urbane adversary adduces secondary qualities, and perhaps
even deigns to mention his favorite red, yellow, or blue, quite
as if these matters contained no intricacies worthy of his closer
attention, and none that could throw any light on his own pro-

1 1 have to give this on the authority of Bering, aa I do not readily find the state-
ment in Helmholtz's writings.

QUALITY AND QUANTITY 331

found views as to their 'subjective' nature. Indeed our adver-
saries, particularly of the present generation, are so emancipated
from allegiance to anything that can be called empirical, look to
it so little for guidance, and argue on grounds so a priori, that they
can only expect some day or other to be punished for indolence and
frivolity. The blessed Absolute will certainly strike out at them
with a burst of novelties : for the Absolute is a powder-magazine
as well as a precious pot of slumbering ineffables. Now the sec-
ondary qualities present interrelations both fixed and intelligible,
so that those persons who seriously study them begin to see that
they form a system like the systems discovered in mathematics,
and this fact alone, as some one has said, already sets them off
from the purely 'subjective/ individual, and incalculable. Un-
questionably the most comprehensive single treatise on these
interrelations is the remarkable work of Dr. Brentano, "Unter-
suchungen zur Sinnespsychologie." * He dwells primarily on the
relation of similarity between the qualities. Thus in the spectral
series of colors, every hue has a position intrinsic to itself by virtue
of its similarity to the adjacent hues : a particular orange has no
place in the series except just between a certain yellowish red on
the one side and a certain reddish yellow on the other. To assign
it any other position would be like trying to assign to the number
3 a position in the number series between 528 and 529. (I
know that some subjectivists declare with their usual license that
this can be done, but before they stake great hopes on doing it
I urge that they actually try and then report to us from time to
time on progress made. One can place orange in space among the
reddish purples, but it hits back at one because the spatial order
has been made discrepant with the 'intrinsic,' logical color order.)
And the case is the same with grays, with auditory pitches and
timbers, with warmth and cold sensations, to some extent (which
increases with careful study) with tastes and odors : the same
thing is true of all intensities in whatever field of sensation.
Now why, asks Brentano, is this relation of similarity so rigid

1 Leipzig, 1907.

332 ILLUSORY EXPERIENCE

and inexorable? The reply has often been that the question
admits of no answer, that namely this qualitative similarity is
an ultimate, unanalyzable category; the very answer that for
centuries was made regarding the similarities between plant and
animal species and between chemical substances. God made
them so ; voild tout. The modern answer is identical save that,
agreeably to the whim of the day, God is left out. But Dr.
Brentano is something of an empiricist, and knows the empiri-
cal grounds for the maxim that quality is not an ultimate cate-
gory of science. He looks then, but only looks and does not a
priori argue, for some quantitative resolution of these qualita-
tive relations of similarity. And he finds them in profusion. It
happens that ever since the day when Helmholtz made a group of
tuning-forks pronounce the vowels, and he and Stumpf showed that
musical chords can be analyzed, not merely physically but as
purely 'subjective' phenomena, it has been generally conceded
in psychology that a quality may seem as unitary as possible,
and that it may nevertheless be, still merely as a conscious phe-
nomenon, complex. It is admitted in short, though sometimes
grudgingly, that even consciousness requires careful study, like
any other phenomenon, and that a seemingly simple quality may
only need practised scrutiny to be resolved into separate ele-
ments. The merest novice may have an inkling of this if he will
strike the prongs of a fork from the dinner table, press his finger
very lightly thereon, and ask himself carefully whether he feels
a simple quality of roughness or a microtactile succession of im-
pacts. In fact, the analysis of conscious qualities is precisely
like the analysis of chemicals ; if careful study yields an analysis,
the phenomenon was not simple ; if components are not isolated,
the phenomenon may be simple, or it may yet be analyzed by fur-
ther study.

This principle is now endorsed by the most conservative au-
thorities, for the entire range of musical chords, timbers, and every
other variety of auditory quality, except pitches, and including
(with a few dissenting opinions) noises. The principle is- not so

ANALYSIS OF QUALITIES 333

universally admitted in the field of color, and here Brentano
commences his investigation — "Concerning Phenomenal Green."
Is green a simple quality with 'similarities' to yellow and blue,
or is the green composed of two elements, yellow and blue ? In-
trospective reports differ, and the question arises whether this is
because the judgments are based merely on associations (as in
some careless investigations in aesthetics) and are subject to va-
rious 'illusions of judgment,' or whether the divergence arises
because the individuals reporting have varying degrees of skill
at introspective analysis, as in the case of the analysis of a chord
by musical and unmusical individuals. For Brentano the latter
is the case here, and in my opinion he thoroughly substantiates
his opinion l with a wealth of fact and a scrupulous precision of
treatment, which leave nothing to be desired. The conofensusof
opinion among painters, who are here the trained musicians, is
that green is composite, and phenomenally is yellow and blue.
This is not owing to the fact that they often make their greens
by mixing yellow and blue pigments, for on the one hand they very
often make green by mixing yellow and black ; and on the other
hand, while they as often make yellow by mixing orange and
green pigments, they never see in the phenomenon yellow —
orange and green. They make 'fine gray' (S. 153) too by mix-
ing red and green, and another gray by mixing yellow and blue,
yet they never see in the resulting gray these pairs of colors.
Brentano's argument touches every conceivable aspect of the case,
and I must here refer the skeptical reader to the original work.
Green is, in short, an interesting, because disputed, case like
others which are more generally admitted. More persons admit
that they see orange as red and yellow, that purple is red and blue,
that greenish blue is green (or yellow) and blue, than admit that
green is a composite quality. There is, on the other hand, a strik-
ing unanimity of opinion among the practiced as well as the un-
practiced, that white resembles no color, that gray resembles nothing

1 Op. cit., 5-49, 129-158. It may be remarked, by the way, that Dr. Brentano
seems to be quite untainted by realistic heresies (cf. 1).

334 ILLUSORY EXPERIENCE

except white and black, and that red, and yellow, and blue are
phenomenally simple.1 Thus the rudiments of a mathematical
system among the colors begin to emerge, as once analytical
chemistry emerged, by virtue of unanimity among the penetrating
and expert, and in spite of some dissent from those less experi-
enced.2 Brentano brings out many other analyses of color
qualities into their simpler components : but of these I will men-
tion but one. Is it true, he asks, that red and green are as an-
tagonistic as Bering, for instance, declares them to be when he
says that they can never be perceived together at one place?
And if so why is it that olive-green seems to so many persons to
contain a red component? Brentano says much more about it,
but I will leave the question here to the reader's introspection;
admitting for myself that I have always seen red in olive-green and
have never dared to say so.

In the field of audition Professor Brentano takes up an analysis
by Professor Mach 3 of the pitch series, into merely two qualities —
'Dumpf and 'Hell'; such that the various pitches consist solely
of these two qualities combined in varying proportions. Bren-
tano does not fully agree with this, but agrees so far as to admit
the qualities, dull and bright, and that the lower and higher pitch
ranges are always tonally unsaturated by reason of an admix-
ture with one of these qualities, precisely as dark and light colors
are always unsaturated by reason of admixture with black or white
respectively.4 There follows an incisive and careful account of
the matter, based on introspective judgments, which is not wholly
out of joint with the valuable contributions of Stumpf in this same

1 Cf. H. Ebbinghau3 on the 'turning-points' of the color quadrilateral (Grund-
zuge der Psychologic. Leipzig, 1905, 197-198) , also Titchener, E. B., A Textbook
of Psychology. New York, 1909, Part I, 60.

* An interesting paper on the mathematics of this system is K. Zindler's Ueber
raumliche Abbildungen des Continuums der Farbenempfindungen und seine mathe-
matische Behandlung. Zsch. f. Paychol. u. Phyaiol. d. Sinneaorgane, 1899, 20,
225-293.

1 Mach, E., Die Analyse der Empfindungen, 2te Auflagc, Jena, 1900 ; Kapitel
"Die Tonempfindungen," 169-205.

* Brentano, op. tit., 99-125.

ANALYSIS OF QUALITIES 335

field. On these and many other purely introspective empirical
analyses of the qualities Brentano bases his principle of 'com-
pound qualities' (multiple Qualitat),1 the principle, namely,
that the qualities presented in consciousness are generally not
simple, not even 'phenomenally' simple, but are composed of two
or more qualities which careful introspection enables us to appre-
hend as integral parts of the presented quality. This principle
is by no means new, being merely an extension to the entire field
of conscious qualities of the principle of analysis already brought
to approval and acceptance through their work on musical chords
and timbers, by Helmholtz and Stumpf. And I think that an
unbiased survey of the facts adduced by Brentano must convince
any one of the correctness of this extension of qualitative analysis
which is offered in the principle of compound qualities.

Now in recent years the traditional distinction between the
'different senses,' i.e. the modes, has been generally breaking
down. It is at best based on nothing but the gross anatomy of
sense-organs, and has just about that weight of authority which we
concede to the popular 'five senses of man.' Apart from the facts
that the ' sense of touch ' is supplied with at the very least four
different sets of nerves and sense-organs, and the tongue with at
least three types of sense-organ and three distinct cranial nerves,
we have innumerable conscious units, phenomenally considered,
which at first sight seem as simple as colors or musical chords,
but for which we know that the physiological apparatus of pro-
duction involves a variety of different senses.2 The most familiar
case of this is the so-called 'taste' sensations, of which all but
four (sweet, salt, bitter, and sour) involve the cooperation of an
indefinite number of olfactory fibrils, besides, very often, that of
touch, warmth, cold, and pain fibers as well. Thus a taste which
until analyzed seems as simple as the color yellow is yielded by
no less than three of the 'five senses' working in cooperation;
or, in the modern terminology, by the action of the taste, smell,

1 Brentano, op. cit., 159-161.

* This fact is brought out well by Nagel in the essay previously referred to.

336 ILLUSORY EXPERIENCE

tactile, and perhaps the warmth, cold, and pain 'modes' all
together. Almost every familiar touch sensation (smoothness,
wetness, et cetera) is recognized as being similarly compounded
of two or more dermal modes. And the number of such seemingly
simple, yet physiologically complicated, sensations is beyond
reckoning; and were it to come to such conscious entities as
"the feeling of triangularity," which several psychologists have
asserted to be simple, the variety exceeds all bounds. In short
the modal boundaries, perfunctorily taught in many elementary
textbooks, has no meaning at all for a psychology which has
outgrown its swaddling clothes (cf. Nagel, op. cit.). Here too
Brentano comes in with fresh arguments, and notably in the
matter of intensity. Now intensity is admittedly a feature of all
sensations and perceptions, and in so far binds together the dif-
ferent modes. Yet only rather recently has it been shown l that
the intensities of two different qualities even within a mode can
be accurately compared and a judgment pronounced of more and
less. Professor von Kries now says that this is undoubtedly pos-
sible for colors however distinct in quality. Dr. Brentano goes
further, showing that hi such cases as of a very faint odor com-
pared with a loud noise, it is possible to judge that the latter is
more intense than the former. How far practiced introspection
may enable us to go toward the refinement of such judgments,
nobody as yet may know. There are also other communities
between the different modes : "A sound that approaches a
noise in character we ... pronounce a less saturated tone sen-
sation than another which is less like a noise. In the field of
taste, indeed, Aristotle correctly remarks that sweet is related
to bitter as is a brighter to a darker color. And similarly sev-
eral persons whom I have asked have definitely pronounced the
cool sensation of a breeze blowing on the hand as brighter than
the feeling of a warm breath. A sensation of coolness com-
pared with that of sweetness or with the odor of a lily, is as un-

1 Cf. von Kries, J., Nagel's Handbuch der Physiologie des Menschen, Braun-
schweig, 1904, 3, 256-267.

ANALYSIS OF QUALITIES 337

saturated as white compared with a color in the narrower sense,
or as a hiss or other noise compared with a vibrant tone.1 " By
such examples this author shows not only that the infinity of
qualities within the mode consists, phenomenally, of combina-
tions of elementary qualities, but also that the modes themselves
merge into one another and present, in part at least, common
elements such as intensity, saturation, brightness, and so forth.
Lastly I will mention Professor Brentano's thesis that no quality
which has intrinsically a position between two other qualities can be
simple. It must rather be a compound of the qualities between
which it so obviously lies. For in what, otherwise, should the two
similarities, this necessary betweenness, consist ? Here he takes,
as unequivocally as possible, the accepted position of science, that
qualitative 'similarity' is never an ultimate category.

However one may agree or disagree here with some of the par-
ticular cases, enough is proved, and I think very amply proved,
to show that nearly if not quite all of the so-called secondary
qualities are, taken merely as phenomena, complex; that careful
introspection enables us to analyze out many, more elementary
qualities; and that, furthermore, we cannot as yet seethe limits
to the possibilities of such analytical procedure. Such a psy-
chological 'atomism/ for this is what the whole matter points
to, has been shown by other writers to be the course which the
facts are clearly leading psychology into. Professor Miinster-
berg is one who has urged this.2 "Are these sensations," he writes,
" the ultimate elements of our consciousness, or is that which we
call a blue or a hot sensation, a sweet taste, a tone C, a muscle
sensation or a pain sensation itself a complex affair which consists
of more elementary parts : in short, have we in the mind ultimate
elements which are simpler than the sensations? It is the in-
quiry for a radical psychological atomism." (pp. 4-5). "The
psychological fact which stands immediately in the foreground
of such considerations is the fact of the similarity of the

1 Brentano, op. tit., 80.

1 Miinsterberg, H., Psychological Atomism. Psychol. Rev., 1900, 7, 1-17.
z

338 .ILLUSORY EXPERIENCE

sensations. . . . Similarity from the point of view of descrip-
tion is community of parts : . . . The logical conclusion by
analogy is that two sensations also are similar to each other only
when they contain various component parts of which some are
common to both. These parts are of course not sensations but
inexperienceable factors like the atoms, and as we do not know a
sensation which is not in some way similar to some other one, we
can say that no known sensation is an ultimate element" (p. 11).
Thus for Professor Miinsterberg the ultimate elements will neces-
sarily be 'absolutely dissimilar' from one another, since so long
as any similarity subsisted, it would indicate the presence of some
yet similar common ingredient. The same should indeed seem
to follow from Professor Brentano's principle that no quality is
simple which (by similarity) lies between two others. Both
writers agree that all the traditional modal "demarcation lines
which existed for the sensations have now disappeared. . . . The
similarity between smell and taste, or between touch and muscle
sensation, and so on, appears then not different from the similarity
of two tones" (p. 13). Professor Miinsterberg adds that only the
ultimate elements, "the psychic-atoms can rightly be correlated
with the physiological units" (p. 16), meaning to suggest, clearly,
something like cortical cells or synapses.

This is a very promising program, but one point needs fur-
ther examination. We seem to come out with a set of 'utterly
dissimilar' psychic elements, and we seem to have no clue as
to how many of these we are likely to find; it will apparently
not be less than the number of cortical elements. This must be,
of course, as it empirically is found to be; and yet an analysis
with so stated a limit is so different from all other scientific analyses,
that one is prompted to look through the argument once again.
Besides, scientific analysis arrives at similars and not at differents.
Will it pay us to undertake the analysis at all if at the end the
number of irreducible elements is to be not less, and perhaps far
more, than the number of brain-cells or synapses ? Will that be a
kind of science to which the human mind shall advisedly address

PSYCHOLOGICAL ATOMISM 339

itself ? We have already seen that introspective analysis is pre-
cisely comparable to chemical analysis, and Professor Miinster-
berg's program promises as its goal very much that thing which
chemistry had attained when it had reduced all substances to
some seventy not further reducible types of atom. But we now
see that this was not to be the goal of chemistry ; the periodic law,
largely, dissolved these 'atoms.' A further comparison with
chemistry shows the miscalculation in this psychological program.
And this lies hi the neglect of an important aspect of analysis.

If it is true that a quality which lies between two others must
have a common ingredient with them, it does not necessarily
follow that for the original three qualities the total number of
ingredients is now four (one common and three proprial). Such
cases occur, but they are utterly atypical. Typically the three
qualities are at first found to consist of two ingredients only,
combined in different numerical proportions ; and finally of only
one substantial ingredient, such as one kind of atom, variously
organized, such as three sizes of molecule. And science has the
best empirical evidence for accounting no analysis complete until
it has reduced all qualitative differences to different arrangements
of elements which are all alike in quality. (And 'quality' here
is presently seen no longer to have a meaning.) In physics and
chemistry these arrangements are confessedly numerical values
organized in time and space. Now it is here, probably, that Pro-
fessor Miinsterberg finds the rub, for he denies (for reasons of a
somewhat 'metaphysical character) that psychic atoms exist in tune
and space, or can be organized therein. It should seem to be
for this reason that he sets so singular a goal to psychological analy-
sis. Now we can for the moment grant him the timeless aspatiality
of psychic atoms, since a merely numerical, or, as I should per-
fer to say, logical principle of organization is quite enough for
our purpose; — of which the number system itself is an instance.
This principle alone will give, from one type of element, all the
variety which we require, as any one shall see who will trouble to
enumerate all the positive whole numbers, to say nothing of the

340 ILLUSORY EXPERIENCE

negative, fractional, irrational, and 'unreal' numbers. There
seem, then, to be no reasons of a general nature why we should not
hope to reduce by analysis all of the conscious qualities to different
forms of organization of one sort of element. So much for theory ;
how are the concrete facts ?

They are most patently that psychic entities in general (and
not merely elements) are organized into higher units precisely
in time and space. The ' Gestaltqualitaten ' are precisely such
higher units : any rapidly phrased sequence of a few tones yields
a form-quality which psychologists have repeatedly declared to be
a new, unique, independent, and unanalyzable quality ; and yet it
is a thing of which the structure remains conspicuous. New and
unique it may be, but*it is neither independent nor unanalyza-
ble ; indeed, the conscious analysis cannot be inhibited except by
the most purblind votary of preconceptions. Certainly there is
novelty about a new form of organization ; water is, indeed, more
than oxygen and hydrogen, it is these with organization added.
And this is precisely what the analyst says, — water is hydrogen
and oxygen and organization. So with the form-qualities. Their
number, too, is without end in all the fields of sensation where
succession can be perceived, and I make bold to affirm that none
can deny that here the psychic components are organized in time
(just like 'ether' vibrations) except after an a priori meta-
physical preamble which is deliberately calculated to obscure
the salient empirical fact.

Now we have form-qualities in space as well as in time. The
late Professor James has dwelt on the integral aspect of such
'feelings' as that of triangularity, rotundity, squareness, and one
quite grants this as one grants the integral aspect of water. In-
deed I should add millions more, for I discover a similar integral
aspect to the comic and the tragic, to the picturesque, the blithe,
the wholesome, the inquisitorial, and the horrific. (While the un-
analyzed aspects of the canonic-pretentious, the idiotic-fallacious,
the obscurantic, and the conscious-deceptive are units to keep well
in mind so that one may recognize and name them when one

FORM-QUALITIES 341

meets them — whether in one's self or in another.) These all,
however, have their disintegral aspects as well; they are clearly
complexes organized temporally, spatially, or logically, or in more
than one of these ways at once. The beauty, however, of the purely
temporal or spatial form-qualities is that they reveal more clearly
than any instance ever shown in chemistry, that the peculiar
flavor of the whole does not devour and supersede the distinct-
ness of the parts, or that the relatedness of the parts is not repug-
nant to the unity of the whole. If one finds it at all within one's
powers to believe that the water which one drinks is oxygen and
hydrogen spatially combined, how much more should one see that
the Graal Motiv is a temporal succession of tones, and that the
triangle is perceptions of line and angle spatially ordered : for the
water does appear to supersede its gaseous components, whereas
in the form-qualities the synthesis and analysis are to be observed
simultaneous and amicable. Once more I affirm that one cannot
deny the spatiality of the organization of the conscious elements
of a triangle, except one close one's empirical eye-lids in sleep.

But these cases seem remote from the secondary qualities,
our theme. I mention them merely because their temporal and
spatial organization is so indubitable. Yet from these form-quali-
ties we can pass directly down to the qualities where the formal ele-
ment almost or quite defies introspection, and we shall find
examples at every step. A paradigm than which nothing clearer
can be desired is a series of light touch stimuli given at an ever
increasing rate. The single tap is called a conscious (secondary)
quality. A pair in slow succession is already a form-quality, with
all the vaunted charms of novelty, uniqueness, and what not ; yet
I will not believe that even amid such delights any one can for
an instant find there anything more than two taps plus temporal
organization (I grant that this last consists of more than mere
twoness). A pair given in more rapid succession is another
form-quality, 'entirely different7 from the first, as we may read
in many of the textbooks. It is not entirely different, but it is
different, and by just as much as the difference of temporal or-

342 ILLUSORY EXPERIENCE

ganization has altered it. Here too the unity of the whole does not
infringe on the distinctness of the parts. As the taps are given
in faster and faster succession, however, there comes a time when
encroachment sets in (at somewhere around 3 taps per sec). The
integral aspect is approaching that to which we give the name
roughness (but is not yet roughness), and the attention is drawn
ever and anon from this to the other aspect, which now consists
in an attempt to count the taps and to articulate the number
with the tongue. The two aspects have begun to behave precisely
like the two aspects of water — either H2O or water. The in-
compatibility of the two aspects depends, however, on no feature
of the new form-quality, but on a definitely physiological defect
of the tune-sense which has begun to be flustered and calls in the
tongue to its aid and so spasmodically wins to itself the balance
of attention. This is precisely comparable to those time-errors
mentioned in an earlier part of this paper. Even here favor-
able moments arise when the integral and disintegral aspects are
apprehended together and as not antagonistic. As now the rate
of tapping increases the unified aspect approaches the (form-)
quality of roughness, and more and more tends to usurp the at-
tention; while the succession-of-parts aspect becomes vaguer.
Articulation is outstripped, then the count is lost, then a sympa-
thetic, voluntary, inner rhythm (of the vocal cords?) which re-
placed lingual articulation falters and fails, leaving a bare and non-
participating awareness of succession. This last persists up to
very much higher rates, dying down only at about 600 taps per
second. Meanwhile the roughness has become a distinct quality
in its own right, so that some psychologists describe it as subsist-
ing with, but 'utterly independent' of, the awareness of succes-
sion. They are the same who find water utterly independent or
hydrogen and oxygen; while the truth is merely that with the
increasing rate of the taps comes increasing (logical and tem-
poral) complexity of relations, to meet which the attention
(with its time-sense and other auxiliary implements), no longer
able to survey the whole intricacy, covers such parts as it best

SECONDARY QUALITIES 343

can. Again (to resort to my all but evaporated simile) this is
like the many properties of water, all of which are deducible from,
and logically bound up in, the properties of oxygen and hydrogen
and their form of combination; yet so intricate is the whole
manifold, that the scientist thinks himself fortunate if in a life-
time he can make a survey of even the vapor-pressure properties
of oxygen or the role played by water in dropsy of the mammalian
heart. This is what there is, and all that there is, in the current
semi-mystical rubbish about the incorruptibility of wholes into
their parts. A whole is different from and independent of its
parts for precisely those persons who find the word 'independent'
so censurable when uttered by the realist.

Thus in the experience of tactual roughness the properties of
the ordered whole gradually supersede hi attention the properties
of the parts ordered, not because the whole, as the rate of tapping
increases, is coming to be anything other than the ordered sum of
its parts, but because the quicker succession soon eludes the sense
of tune, and so leaves other features of the succession (which more-
over is itself now a more complicated thing, per tune unit if not
also otherwise) to occupy the attention. This is merely to say,
what we have found all along, that the faculty of attentive intro-
spection (and notably in so far as this relies on the time-sense)
has its distinct limitations. Stated with less concreteness but
with more logical precision, the case is this. Even a slow suc-
cession of taps (and even were each of these logically simple like
a geometrical point) is already a logical system of no mean com-
plexity ; as one sees from the number of derivatives discovered
by differential calculus where a rate of succession is involved. It
has therefore distinct parts which I have in deference to a bad
tradition called 'aspects.' Now the case of the rather slow
tapping shows us that the series is at one and the same time both
a division of time and a quality, roughness; by which I do not
mean that the quality is anything over and above, or added to
the pure succession of taps. It is some intrinsic feature of that
succession, logically bound up in it. The quality part may be

344 ILLUSORY EXPERIENCE

related to the time-division part as the mere arithmetical number
of railroad ties to the mile is to the spatial distance between the
ties : these two values are simple functions of each other, of course,
and yet mathematically not identical. Indeed, the actual experi-
ence of the taps strongly suggests that the quality is the number
of taps per unit of time (i.e. their density) as distinguished from
the (perceived) time interstices between them ; for as this time-
perception wanes, the (to be sure uncounted, which is a still dif-
ferent matter) number-quality, if I may so call it, waxes. To
show that this is by no means whimsical hair-splitting, I may
remind the reader that in railroad construction one speaks of a
cheap or expensive road-bed, quite as if these were 'qualities'
of the bed. Now this quality of expense is held to vary directly
with the number of ties to the mile (for one of the factors) ; while
the very different quality of safety of traffic is held to vary directly
(for one factor) with the nearness of the ties to one another. And
even to a layman it would seem less direct to state the safety of
traffic in terms of the number of ties to the mile, or the expense
of construction in terms of the spatial vacancies between the ties,
although either could be done and by some accountants doubt-
less is. Similar niceties abound in the computation of traffic-
density, train-load, and so forth.1

Again a rate of motion is a rate of motion, yet it has, not by way
of additions but by way of inclusion and inalienably, first, second
and n derivatives of space with regard to time : these are parts of
it just as number of taps per time unit and the divided time inter-
vals are parts of the tap-sequence. Yet as intimately as the rate
of motion and its first derivative are related, so that to loose
thinking "one is merely a different aspect of the other," the rate
of motion may be very great while the first derivative is zero;

1 Again, to borrow a simile from Professor Royce, how unrelated is the shiftless or
penurious habit of shaving every other day, to the pious practice of shaving im-
maculately every Sabbath : surely as different as the quality of tactual roughness
is from the quality of tactual slowness. Yet the oft-demonstrated incompatibility
of shiftlessness and piety rests on the mere numerical triviality that 3 times 2 is an
even number, while 7 times 1 is odd. The parallel is imperfect.

QUALITY AND DENSITY 345

the body is moving rapidly, but accelerating not at all. And
when this little mathematical scheme of relations is found to pre-
vail hi concrete affairs it often determines extreme qualitative
contrasts. A man may be very rich (a form-quality surely)
and happy (another and very different form-quality), yet the being-
rich depends on his rate of income while the being-happy or un-
happy depends on the first derivative of the rate of income — the
rate of accumulation; for however rich a man is he sees and feels
his fortunes bright if his rate of income is increasing (i.e. if the
first derivative is positive) ; but he sees them dwindling and feels
that ruin is only a matter of time if the first derivative of his rate
of income is negative.

I hold it then to be no academic act of logic-chopping if I say
that the experience of the tactual rhythm shows that tune-division
and number ('AnzahP) of taps per tune-unit are distinct though
inseparable parts of this sequence, and that this latter feature is
just the quality of roughness. At some rates the attention may
be given to both features simultaneously or to either at will;
at other rates the powers of attention (introspection) prove inade-
quate and only the quality can be apprehended. This reveals a
limitation of the power of introspection, and more specifically
of the time-sense, but not any cleavage between the inseparable
features of time-division and tap-frequency.

Now as the rate of tapping becomes still faster the quality of
roughness changes continuously toward the quality of smooth-
ness (just as in the spectrum red changes to yellow), until even-
tually it has run its gamut and becomes the steady quality of con-
tinuous touch; which under other complications is a component
of the form-quality solidity. The precisely parallel experience is
to be had hi the field of audition where a slow sequence of faint
thuds or hisses passes over into a musical pitch. Here too there
is a rate (best gotten with a siren of governable speed) at which the
sound sequence and pitch are experienced together, as distinct
but inseparable. Here too careful scrutiny shows that the
marked-off time-intervals are tune perceptions, while the repeated

346 ILLUSORY EXPERIENCE

hisses are the quality of pitch : the two are just as distinct but
just as inseparable parts of the given succession as are the first
and second derivatives of a rate of motion. In perception they
are separable (as in mathematics distinguishable) just as in per-
ception the form is separable from the size of a concrete object : —
so specific is the action of the nervous system in selecting what
features of a complex shall enter consciousness. The same ex-
perience is had again when the beats between two tones (nearly
alike, but steadily departing from each other in pitch) gradually
pass over (with a disagreeable transitional stage of dissonant
roughness) into a third pitch, the difference-tone. Anent which
we have some amusing speculations, called introspections, by
persons whose logic has vitiated their introspection. These per-
sons say that because the beats and the difference-tone can at a
certain rate be apprehended simultaneously, the two "must be
utterly independent" entities — a point which we have already
dealt with. The parallel experience can be had in vision where
an intermittent color stimulus is seen with increasing rate of
flicker. At first it is succession without the flicker quality, then
succession and a form-quality of flicker, then less obvious succes-
sion and more distinct (form-) quality, then the quality of
(rough) flicker, then the quality of smooth flicker, then the end.
This series is run through rapidly, since flicker entirely disappears
at 60 to 70 alternations per second. The plain lesson of all these
cases is that what we call secondary qualities are in truth form-
qualities, simple psychic entities in temporal organization (rate
of sequence) and we are tempted to view them as qualities instead
of form-qualities because for all but the most absurdly slow rates
of succession the time-sense is inadequate to its task. The succes-
sive 'aspect' is unperceived while the multitude 'aspect' (it
might be by mere summation) gains on the attention. The time-
sense throws over its duty so preposterously early that it is no
wonder that for the upper range of pitches and for all the colors
(which we know are 'correlated with/ and which I am aiming
to show are identical with, extraordinarily rapid successions) every

QUALITY AND DENSITY 347

trace whatsoever of a form-quality should elude the introspec-
tion.

We are now in a position to sum up what I fear will have been
a tedious and apparently aimless array of evidence. The prevalent
form of the specific energy theory, or the view that the secondary
qualities present ultimate unresolvable 'qualitative differences'
(like the immutable species deposited by the hand of God in
the Garden of Eden), has utterly broken down. It is actually the
view that the qualities are independent, unanalyzable, psychic
substances, and there is no good evidence that Johannes Miiller
ever intended his theory to take this shape. These qualities,
considered merely as phenomena, reveal to introspection compound
structure (musical chords, timbres, colors like blue-green, orange,
violet, sensations such as roughness), 'similarities' (odors), and
other affinities and repugnancies which prove that these quali-
ties are not so independent as was supposed, but are at least
related as different species are related in a genus. Further, the
field of consciousness presents innumerable instances of psychic
components organized into higher unities by means of spatial
or temporal relations (form-qualities). Indeed, the entire field
of consciousness presents nothing but such complex entities in which
the principle of organization is either spatial, temporal, or logi-
cal, and a class reputed to stand quite apart, of entities alleged
to be simple and independent — the 'secondary qualities.' These
complex entities have indeed a unified 'aspect,' on which alone
some writers are prone to dwell, and otherwise they were not units
at all ; but they each and every one, as I have tried to bring out,
reveal unmistakably the unity of the whole undispelled by the
relatedness of the parts, and the ordering of the parts constitutive
of the unity of the whole, as even no example of chemistry or
physics so reveals. To any but the most purblind gaze they
utterly scatter and quench the neo-Hegelian imbecilities about
'vital,' 'organic,' unanalyzable, and altogether ineffable 'wholes.'
Alone the secondary qualities maintain their hereditary Sonder-
stellung. But we have found a breach in their wall and now know

348 ILLUSORY EXPERIENCE

how defenseless is their position behind it. Any form-quality of
which the principle of organization is time is, like any other con-
tent, dependent on the mechanism of perception for its entry into
consciousness. And when tune is the formal principle, this form, if
perceived at all, must be perceived by aid of the time sense (of which
the physiological mechanism is so far not in the least understood) ;
and this tune-sense is utterly inadequate to perceive the form-
qualities when the tune divisions therein involved are very small.

One hears that the time-sense is most acute: and the most
favorable case that can be cited is that of bare succession. A
single pair of auditory stimuli can be perceived as two, even when
they succeed each other by two thousandths of a second. But
this is not in truth a time-perception ; it is one of mere 'twoness,'
i.e. bare numerical as opposed to a truly temporal discrimination.
And if the stimuli continue at anything like this rate of succes-
sion, the conscious result becomes at once a mere burred sensa-
tion akin to the quality of roughness. The time interval involved
is not apprehended (as time). Tactual roughness can still be ap-
prehended when the stimuli are as frequent as 600 per second, and
this, again, is cited as a feat of the time-sense. But such it is not.
There is no more consciousness of time involved in the quality of
roughness than in that of smoothness or in the quality red, as the
most casual introspection proves. What is interesting about rough-
ness is that its quality varies so immediately with the number] of
stimuli given hi a time-unit that it shows conclusively that the
nerve is able to carry an impulse of this same frequency number,
and that the roughness quality is precisely this frequency magni-
tude and with the time element specifically omitted from con-
sciousness. Just this, now, is what I contend is the case with all
of the secondary qualities : they are all /orw-qualities in which the
temporal subdivisions are so small that the time-sense cannot dis-
criminate them, whereas the frequency magnitude or the density
still remains perceivable : and density is different from time, since
we have it in spatial and even in mathematical manifolds.

Now if the secondary qualities are all such densities, it must be

QUALITY AND DENSITY 349

that, however inadequate the time-sense, the nerve-fibril is at
least able to carry these densities or frequency magnitudes to
the brain (even those of the highest frequencies, which are the
violet end of the spectrum). And this is precisely what physi-
ology is now showing to be the case. Rayleigh and More have
proved that the auditory nerve carries sound-frequencies, up to
512, if not 1024, per second. And Meisling has adduced facts and
arguments of great weight to show that the visual cones are electro-
magnetic resonators, and that the optic nerve must carry impulses
of a frequency proportioned to that of waves of light. Every-
where, moreover, physiology, as distinct from sense physiology,
is finding in the phenomena of 'Treppe/ refractory phase, el
ccetera, that the nerve current shows oscillations which can be
measured only in a few thousandths of a second. If physiology
has been late in discovering this, it is because such frequencies,
to say nothing of higher ones, could by no possibility have been dis-
covered prior to certain very recent improvements of instrumental
technique, and specially the introduction of the oscillograph or
string-galvanometer. I hold it, then, to be a view which is amply
supported by facts, that the secondary qualities, instead of being un-
analyzable psychological elements are all form-qualities of which the I
time-sense is inadequate to perceiving the form, while the density is
perceived for very high frequencies by a process which is perhaps]
related to physiological summation. It is to be noted that the]
secondary qualities seem to engage the attention roughly in propor-
tion to their density, as from our present view we should be led to
expect. Tactual roughness is readily driven from attention (inhib-
ited) by a sound of moderate pitch, this by a higher pitch, and so
on up. But a color of anything like the same intensity, when pre-
sented, makes it difficult to attend to a sound.1

1 Of course all this is true only when the comparisons are carefully made. Atten-
tion factors such as ' Bewusstseinslage,' difference of intensity between the com-
pared qualities, fatigue, and many other factors will obscure the point. I am speak-
ing, too, of cases where the time element is not present, and excluding, therefore,
such temporal form-qualities as rough flicker et ccetera. Where the temporal
organization begins to be perceived, it often strikingly commands the attention.

350 ILLUSORY EXPERIENCE

Our argument comes out to a psychological atomism which is
substantially like that proposed by Spencer.1 "Although the
individual sensations and emotions, real or ideal, of which con-
sciousness is built up, appear to be severally simple, homogeneous,
unanalyzable, or of inscrutable natures, yet they are not so. There
is at least one kind of feeling [musical sound] which, as ordinarily
experienced, seems elementary, that is demonstrably not ele-
mentary. And after resolving it into its proximate components,
we can scarcely help suspecting that other apparently-elementary
feelings are also compound, and may have proximate compo-
nents like those which we can in this one instance identify " (pp.
148-149). There follows an account of the beats going over
into a difference-tone, and so forth. Now "if the different sen-
sations known as sounds are built out of a common unit, is it not to
be rationally inferred that so likewise are the different sensations
known as tastes, and the different sensations known as odors,
and the different sensations known as colors? Nay, shall we
not regard it as probable that there is a unit common to all these
strongly-contrasted classes of sensations? If the unlikenesses
among the sensations of each class may be due to unlikenesses
among the modes of aggregation of a unit of consciousness com-
mon to them all; so, too, may the much greater unlikenesses
between the sensations of each class and those of other classes.
There may be a single primordial element of consciousness, and
the countless kinds of consciousness may be produced by the
compounding of this element with itself and the recompounding
of its compounds with one another in higher and higher degrees :
so producing increased multiplicity, variety, and complexity"
(p. 150).

"Have we any clue to this primordial element? I think we
have. That simple mental impression which proves to be the
unit of composition of the sensation of musical tone, is allied to
certain other simple mental impressions differently originated.

1 Spencer, H., The Principles of Psychology. Third Edition, I, Part II, Chap.
I, (New York, 1892, i, 145-162).

PSYCHIC ATOMISM OF SPENCER 351

The subjective effect produced by a crack or noise that has no
appreciable duration, is little else than a nervous shock " (p.
150). Quite so, and I think that the experience which I before
called the still small voice of psychology, somewhat approxi-
mates this primitive nervous shock. Spencer gives many further
illustrations. "The fact that sudden brief disturbances thus
set up by different stimuli through different sets of nerves, cause
feelings scarcely distinguishable in quality," he continues, "will
not appear strange when we recollect that distinguishableness
of feeling implies appreciable duration ; and that when the dura-
tion is greatly abridged, nothing more is known than that some
mental change has occurred and ceased. ... It is possible, then
— may we not even say probable — that something of the same
order as that which we call a nervous shock is the ultimate unit
of consciousness; and that all the unlikenesses among our feel-
ings result from unlike modes of integration of this ultimate
unit" (p. 151).

In view of the time at which this was written (about 1855),
the speculation seems astonishingly bold. Even more so, to me,
seem certain remarks (pp. 152-153) on the oscillatory nature of the
nervous impulse, and I should much like to know what physi-
ological investigations were then available, which justified the
belief in nervous frequencies at anything like the rate of sound
or light. Nevertheless subsequent investigations have remark-
ably confirmed Spencer's view (I had almost said, intuition), and
it seems to me that to-day this atomistic theory of consciousness,
hardly modified from the form which Spencer gave it, must be
looked on as a very shortly to be demonstrated fact. Most sub-
stantially supported by empirical investigations it already is.
For myself I quite adhere to this view, while dissenting from Spen-
cer's further remarks as to the nature of the primordial conscious
unit, and from other features of his philosophy of mind. In
weighing this theory one should never forget the phenomenon of
roughness, which (whatsoever the sense-organ that originates it)
is to introspection a quality, and is at the same time nothing else

352 ILLUSORY EXPERIENCE

than the density of a succession of conscious shocks, which seem
individually to be of quite infral-modal primitiveness.

This completes our argument. And now I can reply to the
anti-realist's questions: How can realism pretend to assert the
reality of the color, sound, and so forth which are vividly present
in the dreams of a person sleeping, it may be, in a box no bigger
than his coffin? Realism, I say, can assert this because the
nervous system, even when unstimulated from without, is able to
generate within itself nerve-currents of those frequencies whose
density factor is the same as in ordinary peripheral stimulation.
I have not said, be it noted, that the density factor of the nervous
impulse is the secondary quality: it is the density of the series
of some relatively primitive sensation which is the secondary
quality ; and nerve impulses may also have such a density, as also
may Hertzian waves and many other things. The case, then, of
sensory hallucinations, whether due to the 'inadequate' stimu-
lation of nerves or to so-called central excitation, is entirely com-
parable to the cases of illusion which we considered in the earliest
sections of this paper.1 In this connection I should not quite
say with Professor Alexander 2 that, "The illusory character of the
appearance is the defect of our quality. With an organ adapted
to see red we can see only red, no matter how the organ is set
a-working." The hallucinatory quality needs to be explained
more specifically, and this, I think, we have done. We could,
of course, be more specific yet and suggest that such sensory
resonators as the retinal cones would resonate in their own
period to the electricity released by any of the various metabolic
processes going on in their vicinity (as a brass resonator will
sound its own tone in response to any tap) : we could suggest that
the cortical synapses have much to do with the periodicity of

1 If some un-serious objector wishes to know, then, whether these qualities may
on occasion be surprised by an outsider as they flit through some patient's skull, I
will say that it is here precisely as with the same nerve-oscillations when they are
normally instituted. One studies their densities by such means as one finds avail-
able; and these are commonly not the unaided eye or ear.

* Alexander, S., Proc. of the Aristotelian Soc., 1910, N. S., 10, 10.

DENSITY AND QUALITY 353

nerve-impulses : or that nerve-impulses of various periods com-
ing together in the central nervous system must often produce
an impulse of a new and perhaps higher frequency. But specu-
lation on the minuter details ought to wait on investigation. Cer-
tain it is that the nervous impulse at large is oscillatory ; and that
the oscillations do sometimes (without external aid) become very
rapid, is no ground for astonishment.

I must here forestall an erroneous inference that will almost
certainly be made from the foregoing paragraph : this is, that
hallucinatory and other consciousness is in the skull. Quite on
the contrary, consciousness, whenever localized at all (as it by no
means always is) in space, is not in the skull, but is 'out there'
precisely wherever it appears to be. This is, for me at least, one
of the cardinal principles of realism, and a realist would say with
Berkeley that "the rose is really red" and so forth, just as it appears
to be.1 The idea that consciousness is within the skull, current as
it is, has arisen from the obvious connection between modifica-
tions of the nervous system and changes in consciousness. But
this connection can be in other ways than that of a spatial inclu-
sion of consciousness by the nervous system. Suppose, for in-
stance, that the latter is like a search-light which, by playing over
a landscape and illuminating now this object and now that, thus
defines a new collection of objects all of which are integral parts
of the landscape (and remain so), although they have now gained
membership in another manifold — the class of all objects on which
the illumination falls. Here, too, there would be a direct connec-
tion between the members of the illuminated class and the move-
ments of the light : as there is between the contents of conscious-
ness and changes of the nervous system. Any class that is formed
from the members of a given manifold by some selective principle
which is independent2 of the principles which have organized the

I 1 cannot here treat such a point with the thoroughness that it deserves. I
have, however, discussed it at length in a volume, "The Concept of Consciousness,"
which was completed some time since and which will appear, I hope, very shortly.

1 See Perry's essay in this volume, The Importance of the Notion of Independ-
ence.

2A

354 ILLUSORY EXPERIENCE

manifold may be called a cross-section. And such a thing is con-
sciousness or mind, — a cross-section of the universe, selected
by the nervous system. The elements or parts of the universe
selected, and thus included in the class mind, are all elements or
parts to which the nervous system makes a specific response. It
responds thus specifically to a spatial object if it brings the body
to touch that object, to point toward it, to copy it, and so forth.
It responds to a secondary quality which is 'on' a particular
object by, firstly, a similar (and physiologically very complicated)
response to this special color and no other. This last is effected,
the facts seem to show, by the nerves being capable of carrying
a nervous impulse of the same frequency as the vibrations which are
sent through the intervening space by that color. If the nervous
system can pick these up and transmit them, it can specifically
respond to them : otherwise not. This puts the colors or quali-
ties into the nervous system neither less nor more than the fact of
ether or air vibrations of the same period or density existing all
through the intervening space puts these qualities into that in-
tervening space. We are little tempted to believe that the color
of a flower fills all the space between the flower and the eye : and
neither less nor more does it fill, or enter into, the peripheral
nerves and skull. The entity responded to is the color out there,
two factors which involve two factors of response; but that
color out there is the thing in consciousness selected for such in-
clusion by the nervous system's specific response. Consciousness
is, then, out there wherever the things specifically responded to are.
Now hi cases of hallucination, indeed, the colors, shapes, and po-
sitions responded to are not in 'real space/ as we are accustomed
to say ; but they are in a space which is in all respects comparable
to mirror space, and which is just as objective. Of their reality
I shall say more in connection with thinghood and contradic-
tion.

But one more point here. In calling the primitive entity whose
density constitutes a secondary quality a 'sensation/ I meant
in no wise a thing within the skull, nor a thing at all mental or

SPECIFIC RESPONSE 355

subjective in substance.1 To introspection itself this simple
entity is as objective as anything else ; it is simpler, truly, than a
concrete object like a shell or fossil, and so may seem more ab-
stract. But it has all of the objectivity of other abstractions such
as points and numbers (which are as little subjective as the starry
heavens). I have discussed at length in another place 2 the ulti-
mate substance of mental and physical elements, and endeavored
to show that no difference of substance exists between the two
groups. Professor Stout has well called the qualities "secondary
attributes of matter." 3 Mind and matter consist of the same
stuff,4 and the little entity that in aggregates of various densi-
ties constitutes the secondary qualities is not far removed from
the little atom that constitutes physical bodies, and in point of
substance there is no distinction at all. So, it seems to me, we get
an intelligible picture of how the secondary qualities are as ob-
jective as the primary. Their being included or not being in-
cluded in the class of things which we name a consciousness, de-
pends for both alike, on their being specifically responded to by
a nervous system. But consciousness is in no sense at all within
the nervous system.6

4. Illusions of Thought. — In the matter of errors of thought —

1 1 take it that this is in opposition to Mr. Nunn's " realist doctrine which takes
as ultimate data a psychic monad." Proc. of the Aristotelian Soc., 1908, N. S., 8, 149 .

* The Concept of Consciousness — chapters on the Substance of Mind, and the
Substance of Matter.

* Stout, G. F., Proc. of the Aristotelian Soc., 1904, N. S., 4, 146.
4 Cf. Alexander, S., ibidem, 1910, N. S., 10, 16.

8 1 hold this view of the secondary qualities as densities, solely because of its
empirical (and not its realistic) value; but I offer it here on account of its interesting
and favorable bearing on realism. For realism it is not essential that the nervous
impulse should reproduce and transmit to the brain any specific character or property
of the stimulus. And indeed, while my own views are as stated above, I can con-
ceive that realism should successfully dispense with my notion of ' specific response.'
This too is not a theoretical construction in the interests of realism, but my reading
of the empirical facts. Were it merely the former thing, I should feel bound to
define more sharply ' specific response,' which is now the best name I have so far
found for an empirical relation that on further study I hope to be able to describe in
much greater detail.

356 ILLUSORY EXPERIENCE

contradicted opinions, fallacies exposed, disappointed expec-
tations, and outstanding perplexities — we meet again a diffi-
culty which can honestly be urged against realism, one where the
anti-realist challenges and still more inspires the thoughtfulness
and zeal of the realist. Such contradictions subsist, it is quite true,
and if realism took on itself to explain them away realism would,
hi my opinion, be no better than several other obliteration-philo-
sophies which I could name. The task for realism or for any phi-
losophy is not to show that evil is only quintessential good, the
imperfect - perfect, and so on, but to acknowledge the empirical
subsistence of errors and contradictions and to show the sig-
nificance and place of these things in the tissue of the universe.
And precisely here I believe realism achieves one of its most sig-
nal triumphs.

We have already found, and left still outstanding, a charge
against realism which is now just a case in point. Not the illusory
or hallucinatory image as such, it -was rightly said by our oppo-
nent, but such an image when it asserts itself to be or when the
realist asserts it to be a real object, is the crux for realism. And
the sting of this situation would of course be that such a thought
would soon find itself contradicted. The illusion that proclaimed
itself real would soon encounter a higher authority to show up
conclusively its unreality. Or the realist who declared these un-
realities to be real could soon be confronted with the contradictory
empirical evidence that they are unreal. And he certainly could
be. The difficulty in either case is clearly one of contradictory
assertions or opinions — errors of thought — and so we can dis-
cuss these issues together, and in doing so we shall have left
from our previous arguments no other objections outstanding.
The issue includes all illusions of thought, but is more compre-
hensive still, and therefore it logically introduces a new section.

ERROR 357

II

ERROR

1. Images Assert Nothing. — Our first case shall be the above-
mentioned 'crux.' Now it is not true that an image or other con-
tent 'asserts itself' to be real, although I am aware that several
neo-realists and other persons who may be said to evince some
leanings toward realism, have declared this.1 I cannot see with
what right a person asserts that a 'Hruth which the sensation
'reveals' is its own extra-mental existence." This is certainly
revealed, but not by the sensation content itself. Nor can I con-
ceive that any content asserts, for or about itself, truth, reality,
objectivity, or anything else; such content simply is, and any-
thing asserted for or about it is another content and one of a
prepositional nature : this is a thought or opinion and it may or
may not be a true one. The former content simply is, and is
in itself neither true nor false. As Professor Dewey has said,2
"Truth and falsity [and I should add reality and unreality] are
not properties of any experience or thing, in and of itself or in its
first intention" : that is, in its capacity as a bare subsistent. The
fallacy in this notion of the self-qualification of mental contents
lies hi a failure to discriminate the bare content as it subsists,
from (prepositional) assertions about it which may coexist with
it in the mind, either explicitly or subconsciously. The first
content is what logic knows as a term, the second is one or more
propositions. To confuse the two is absurd. But the same
confusion is present wherever we find it said that a sensation or
idea is 'aware of/ 'refers to,' 'points at/ or 'means' its ob-
ject. The content does not refer to its object in any way; it
is a part of the object. Any such assertion about objectivity or
about reality is another (and prepositional) content; it is also

1 E.g. Stout, G. F., Proc. of the Aristotelian Soc., 1904, N. S., 4, 159. Nunn,
T. P., ibidem, 1910, N. S., lo, 201.

» Dewey, J., Mind, 1906, N. S., 15, 305.

358 ILLUSORY EXPERIENCE

supplied by experience, and it may itself be further qualified as to
truth and reality by another prepositional content. It is impor-
tant to discriminate in every case between what has been called
the content and intent, or more accurately between terms in the
mind and propositions in the mind ; and after that, in the words
of Professor Alexander,1 "the intent is for description another
and special part of the content." I do not admit, then, that the
image, whether it is true or illusory, asserts anything about it-
self. Such an image may, however, exist in the mind together
with mutually contradictory propositions about it ; — another case
of errors of thought, which we are to examine.

2. What the Realist Asserts.2 — The realist does also not assert
that an unreal thing (image or whatsoever) is a real thing.
And here the realist insists on the conscientious observance of a
distinction which logic and mathematics have long since known
and scrupulously observed, which even some idealists have perhaps
heard of, but which not one anti-realist nor yet all realists seem
even remotely to appreciate. This is the distinction between real-
ity and being or subsistence. Here is a typical case of 'reality'
confused with being. Professor MacKenzie writes,3 "and so the
new realist seems to be in truth one who is persuaded that things
are just as he apprehends them. The idealist, on the other hand,
maintains that what is directly perceived is never in itself real."
Thus when the realist says that as things are perceived so they are,
the idealist stupidly misunderstands him to say "as things are per-
ceived so they are really" i.e. all perceived things are real things.
But while all perceived things are things, not all perceived things
are real things.4 Stupid as such a confusion is, it will be found to
have been made at some point in every anti-realist argument.
This abuse involves two of Perry's fallacies,6 pseudo-simplicity

I Alexander, S., Proc. of the Aristotelian Soc., 1910, N. S., 10, 2.

I 1 have treated all of the ensuing topics more fully in my Concept of Conscious-
ness.

* Mackenzie, J. S., Mind, 1906, N. S., 15, 318.

4 This is why on a previous page I said that "as they really are " is not a realist's
phrase. « Perry, R. B., /. of Phil., Psychol., etc., 1910, 7, 341, 350.

CONTRADICTION 359

and verbal suggestion: 'reality' is taken to be a simple qualifica-
tion connoting no more than being (as if there were no -unreality
to be accounted for, as there is no non-being) ; and, again, the
fulsome delights experienced by most minds on mention of the
'real ' (especially when exasperated by idealistic futilities and un-
truths), make it a welcome epithet which can be slipped in unpro-
tested anywhere. Small wonder that idealistic arguments which
provide so little of the article should make the most of a pleasant
verbal suggestion: for however troubled the sea of debate, the
comfortable word 'real' gives one a sense of being at home.

Now for realism by no means everything is real ; and I grant
that the name realism tends to confuse persons who have not fol-
lowed the history of the term. For the gist of realism is not to
insist that everything is real, far from it, but to insist that every-
thing that is, is and is as it is. Not a dangerous heresy, this, it
should seem; but it just happens that every form of idealism
has maintained the contrary, has maintained, to use a term of
Professor Dickinson Miller's, some kind of 'false-bottom' the-
ory of the universe. Idealism has either said that since some
things are demonstrably erroneous they are not as, on the same
authority, they are (i.e. erroneous) — and this is the German
way ; or it has said that everything is erroneous, nothing is as,
on the very same authority, it is — the way of Mr. Bradley and
his school. The approaches to idealism are extraordinarily di-
verse, or at least are so represented by their expounders ; but
these latter are in fact primarily engaged with the problem of error,
and either they profess to discover that it is not, and so find them-
selves rather busied with an uneasy conscience; or else they are
obsessed with error everywhere, and consequently fold their hands
in despair. For if I dare profess at all to grasp these anthro-
pological mysteries, the differentia of idealistic philosophers is
their common assignment of first importance to the problem of
knowledge, and this problem has been, from the earliest Greek
times to our own, primarily the problem of error. Realism nei-
ther succumbs to this problem with the non-possumus of the modern

360 ILLUSORY EXPERIENCE

English school, nor tries to explain error away with the solemn
circumstance of the German schools. It meets, if I may venture at
all to speak for it,1 the problem of error by borrowing from logic
and mathematics the well-authenticated distinction between reality
and being. The universe is not all real ; but the universe all is.
Two more distinctions, and we proceed : being is to be distin-
guished not merely from being real, but from being true, and from
being perceived or thought. Realism has a just and proper place
for the functions of perception and thought, but the subjectiv-
ist's contention that "being means always in some sense or other
[sic] being perceived or thought," rests solely, as Perry has demon-
strated,2 on the fallacy of the ego-centric predicament.

3. Contradiction and Being. — The earliest (and very ancient)
'solution' of the problem of error seems to have been that errors
are all matters of opinion, are merely subjective, found only in
consciousness: but that the objective world is error-free, so that
no one need worry lest the universe totter and collapse. This re-
mains to-day the comfortable popular view of the matter. Nor
can I see that the triumphal progress of idealism has brought en-
lightenment. Rather has idealism thrown us back on the origi-
nal difficulty by asserting that everything is subjective, from
which the conclusion must be that error is again ubiquitous.
Yet many idealists, and especially the leather-patch school of
Professor Karl Pearson and his associates, profess not to draw this
conclusion, since they continue to dispose comfortably of error on
the ground of its 'subjectivity.' It is true that Hegel under-
took to treat error much more responsibly, but his solution seems
to have evolved a checkmate of thought and intellection uber-
haupt, so that his followers have no course left to them save to sing
the cradle-song of the Absolute, and, so lulled, to surmount error by
oblivion. Yet error remains a problem for persons who have
kept awake, and one observes that the hated name of realism

1 But of course I offer this, as well as any other lines of argument, as merely my
personal interpretation and defense of realism.
* Perry, R. B., op. cU., 5-14.

CONTRADICTION 361

suffices to arouse even the Hegelians to a disturbed conscious-
ness: "If you won't repose ineffably1 in the Absolute, what are
you going to do with error?"

Now it may be admitted that 'errors' are all of knowledge, or are
in experience; but the important point is another : that all errors are
cases of contradiction or contrariety. One has met error who has ex-
perienced that A is B and that the same A is not B. But the ex-
periencing is not the significant fact, and that all errors are of knowl-
edge is true merely by definition, since contrariety or contradiction is
called ' error ' only when it occurs in some person's field of conscious-
ness.2 The actual problem is the contradiction or contrariety itself :
what is the significance of a universe that holds such things ? And
here, once more, the only solution which appeals in practice to any
one is the ancient one : that only one of two incompatible proposi-
tions is in the universe, the other is 'only subjective.' It is for
this reason that every one of the recent writers against realism
centers his attack about the problem of error or contradiction. I
shall base my remarks on their own assumption that there can be
no contradiction in an objective (or 'real') universe.

This last proposition is always expressed, or tacitly implied,
with an assurance which shows that these gentlemen assign to it
axiomatic validity ; and if they were to be asked how they know
so interesting a fact about the universe, they would infallibly
reply that it is self-evident. On which I should remark that so
far from being self-evident, it is categorically untrue. "Are you,"
say they, "crazy enough to think that you have ever seen an exist-
ing object move both up and down at the same moment? Have
you ever met the round-square, or the A that was at the same time
not-A?" and I reply, "No; are you so crazy as to be able to
think these objects?"

An answer to this is, that while it may not be easy to visualize

1 Vide Sheffer, H. M., Ineffable Philosophies. J. of Phil., Psychol, etc., 1909,
6, 123-129.

1 The reader must here avoid the fallacy of the ego-centric predicament, already
referred to.

362 ILLUSORY EXPERIENCE

or image an A-not-A, yet that one can easily think of an A that
should be, or might be, also not-A. This is a mere defect of
imagery, just as it is an accident of vita brevis that one cannot
enumerate an infinite series although one thinks it readily enough.
Now this is the core of the matter. It is not a "defect of imagery "
which prevents us from visualizing the round-square or the A-not-
A as readily as we visualize a hippogriff, the whale discussing
Jonah, or even a Cook at the north pole. The thought of the
round-square is a propositioned content about a strictly unthink-
able IT : — that it is to be square, and it is to be round, and so
forth. Further than this even thought cannot go : even the
inner eye cannot grasp the square which is also round. One can
think of a point which should move up and move down at the same
moment, but when one images the point, it moves either up or
down, or the two successively. Now this "defect of imagina-
tion" is not a psychological matter at all, but rests on a funda-
mental distinction which symbolic logic and mathematics have
more or less recently made out, between propositions or postu-
lates on the one hand and terms in relation on the other. It is
found here that propositions may subsist together in a set although
they are mutually contradictory, but that such contradictory
propositions can never generate, or be realized in, a system of terms
in relation. Indeed, so harmless, oftentimes, are the contradic-
tions between propositions that the only certain test that proposi-
tions are not contradictory is the discovery of a system of terms in
relation of which the propositions all hold true, or in which they
are exemplified. Thus if one undertook to define a figure such that
it should be the portion of a plane surface included between three
straight lines ; that it should possess four (internal) angles ; that
the sum of these should equal 180 degrees ; and that the sum of its
external angles should equal ten times 180 degrees ; — one would
have no means of discovering whether a contradiction 1 had been

1 1 take no pains here to differentiate between contradiction and contrariety,
because both contain a common and more fundamental element of negation, for
which we so far have no good name, but which is one of the secrets of this difference
between sets of propositions and systems of terms.

CONTRADICTION 363

posited except by appealing to the corresponding system of terms
(plane, lines, et ccet.) which such a set of postulates undertakes to
define. This is of course an appeal neither to physical existents,
nor to the faculty of imagination : nor are propositions more ' sub-
jective' than terms. This distinction between sets of propositions
and systems of terms is of the most profound importance ; it sheds
light, for instance, on analysis and synthesis, the meaning of veri-
fication, concreteness, empiricism, and on the triviality of the ' geo-
metrical method/ or any other, when the prepositional seqence of the
argument swings free from the patient exhibition of terms in their
relations. Now this fact that propositions oppose one another freely
while such opposition or contradiction is never exemplified in a sys-
tem of terms in relation, does not, I admit gladly, tell us all that we
wish to know about contradiction and negation. On the contrary,
it merely opens up a field of study most stimulating to the appe-
tite, and one which at the present juncture I conceive to be the
most promising of any, for both logic and philosophy. But the
considerations just adduced are important in two respects; they
do not purport to explain 'error' (contradiction) away; and
they do show that the problem of contradiction (error) has nothing
whatsoever to do with the problem of knowledge or epistemology.

The dichotomy proposition-term, fundamental as it may be, co-
incides in no wise with the dichotomy mind-matter, subject-object,
nor yet with unreal-real. Conscious images, like physical objects,
are terms in relation, and as the round square or A-not-A is not
found among physical systems of terms, so it is (and for precisely
the same reason) not found among mental systems of terms. What
is in the one case called physically impossible ('unreal') is in the
other case found to be mentally impossible, i.e, unthinkable.
On the other hand, the mind can and does entertain the most con-
tradictory propositions about terms, precisely as physical laws,
which have obviously the nature of propositions, are habitually
in a state of contradiction.

I say 'habitually,' although I know how shocking a heresy
it is to speak of contradiction in any connection with the physical

364 ILLUSORY EXPERIENCE

world. This does not contain, it is true, A's that are not-A's.
Neither does the mental world contain them; and the expression
A-not-A, or round-square, has no meaning at all save as symbol
for a little pair of contradicting propositions. But having dis-
covered this valuable fact, apparently, at some pains, natural
science conceived such an animus against the name 'contradic-
tion' that it devised means for disguising the true cases of (prepo-
sitional) contradiction among natural laws; of which every case
of collision, interference, acceleration and retardation, growth
and decay, equilibrium, et ccetera, et ccetera, is an instance. This
is as follows. A law of motion states always that a physical par-
ticle (or series of them) moves (or shall move) thus and so. If
now two wave-motions are progressing along the same straight line
and toward each other, there will be a moment when a certain
particle will be 'acted on' by both motions at once. The law of
one motion will state that the particle moves up (or shall move
up), while that of the other motion states that it moves down
(or shall move down) at the same moment. Each law opposes
the other, and although the relation is called one of contrariety,
this contrariety is in fact more than contradiction; for if one law
says up, the other not only says not-up, but further specifies down
Logic cannot show a negation more thorough. And while the
impossible-unthinkable does not happen, the result which is char-
acteristic of all contradictions ensues, namely, zero motion ; or hi the
cases of different amounts of energy one motion is reduced toward
zero by as much as the opposing motion has energy to contradict it.
To say that because no energy has been 'lost' there has been no
contradiction is nothing whatsoever to the point. Two laws ot
motion have met in contradiction, and this is precisely the ap-
pointed signal for energy of one sort to be transformed Into energy
of another. Because a third law can be framed (hi terms of trans-
formation, strain, or elasticity) to describe what shall continue
to happen when a contradiction arises, should not blind us to the
fact that it is not until the first two laws do meet hi contradiction
that the third law goes into operation. The case is paralleled in

REALITY 365

the game of chess, where the laws governing the moves of the several
men often come in contradiction with the rule that no two men
shall simultaneously occupy one square. To meet this case a
further law declares that the second comer shall 'take' or anni-
hilate the earlier occupant; and the whole game hinges on such
contradictions. To imagine that hi this way contradiction has
been forestalled, is to do like the fatuous master who commands :
"Stand up, but if you won't stand up, lie down; my orders shall
be obeyed." Not even from the point of view of a static logic
is contradiction hi this way avoided; and modern logic is not
static.1 Contradiction is after all a tame and harmless thing, al-
though a very interesting one. The pretensions of many natural
scientists that they find no contradictions is uncommonly absurd,
because in fact they find little else. That is, all natural motions are
the result of so many partially contradictory laws operating together
that it requires a fabulously clever technique to produce a motion
which is simple or uncontradicted enough to allow any one of the
component laws (or constant functions) to be determined. The
natural scientist may conceive this as he likes, but if our idealist
opponents object to the above considerations I will beg them to
take down the gospel acccording to Kant and read what the latter-
day Immanuel had to say in his "Essay toward the Introduction
of the Concept of Negative Quantities into Natural Science "^
it is one, but only one, of the authorities for what I have ventured
to present above. X

The gist of the whole matter is, that the impossible-unthink-
able never happens anywhere, but that every variety of contra-
diction, contrariety, repugnance, opposition, and negation which
logic itself recognizes is quite as plentifully manifested hi the ob-
jective physical world as it is in the subjective sphere of mind.
A thought, then, which negates another thought is neither more
nor less significant than a physical law which negates another

1 Modern logic might well devise a system of purely logical ' strains ' and ' stresses,'
not in order to conceal the fact of contradiction, but in order to study certain as
yet little understood properties of sets of mutually incompatible postulates.

366 ILLUSORY EXPERIENCE

law. The problem of _error, as that of 'reality,' is in no way
involved in the problem of knowledge.

Now the image in consciousness does not assert anything about
itself, nor does the realist, as it seems to me, assert of it that it
is necessarily real; still less would he assert that all the preposi-
tional contents of consciousness are real. But what I suppose
that realism insists on is that every content, whether term or
proposition, real or unreal, subsists of its own right in the all-inclu-
sive universe of being ; it has being as any mathematical or physi-
cal term or proposition has being; and that this being is not
"subjective in its nature" (a phrase indeed to which in this con-
nection I can attach no meaning). I believe, further, that no
content is 'constituted' by a metaphysical knower or ego, for
I believe that no knower, or ego, such as metaphysics means,
exists. I believe also that no conscious content is 'constituted'
even by the knowing process, in the sense commonly attached by
metaphysics to the word 'constitute.' If the knowing process
ever constitutes its content, it is, I believe, only as a ripple of water
in assuming successive forms may be said to constitute these
forms (if that appears either significant or interesting to any one,
as it does not to me). But particularly meaningless is the asser-
tion of idealism that a mental content is 'subjective' in its sub-
stance or nature, or is constituted by a metaphysical knower
or ego.

As to what reality is, I take no great interest ; nor do most other
persons, for if they had done so, they would have taken more
pains to define it sharply as against the equally and perhaps
even more prevalent unreality. But if challenged, I should hazard
the guess that perhaps reality is some very comprehensive system
of terms in relation. For by reality we seem to mean the thing
most remote from contradiction, and this is with certainty found
only in systems of terms. This would make reality closely related
to what logic knows as 'existence.' If this is correct, probably
all of the terms found in the physical world, also some and possibly
all of the terms found in minds, are real. But all this, so far as

REALITY 367

I know, has been far too little studied. Certain it is that unreality
is no more subjective than reality; for a thing may be objective
and yet unreal, as is commonly asserted of certain numbers and
of some systems of geometry.

4. Contradiction and Realism. — Let us now return to some
arguments of our opponents. In a somewhat over-ingenious
paper l Professor Lovejoy brings up at length the case of hallu-
cination. "While it lasts, the hallucination is for the victim of
it as good 'content' as any other of his perceptions." But "it
was not at the time an object perceived by others. ... In fact,
it turns out that the other percipients at that time perceived as
vacant, or as otherwise occupied, the very space which the hallu-
cinatory object ostensibly occupied . . . and since the testimony
of many witnesses and of the general uniformities of experience
is against his object, the victim of the hallucination, when he re-
covers, proceeds to call his object somehow 'unreal,' and to declare
that the ' content ' that was truly in the space in question was that
beheld by the other percipients. ... Be it observed that the
corrective judgment, through the making of which the concep-
tion of consciousness is generated, predicates ' unreality ' — now
interpreted as 'existence merely for consciousness' — of the ob-
ject of the hallucination at the moment when that object was present"
(p. 595). Professor Lovejoy grants it as possible that the "new
realism . . . does not maintain that the same portion of real space
can be at once both empty and filled," but he does not see "what
account a new realist can consistently give of the status of the hallu-
cinatory object at that time, or of the difference between it and
the coexisting 'real' objects" (p. 596). Now I should not call
the hallucinatory object necessarily 'unreal,' still less an existent
"merely for consciousness," nor need anybody hasten to the con-
clusion that the other object is 'real' since collective hallucina-
tions are also possible. It is to be noted that a "corrective judg-
ment" (i.e. a proposition) in opposition to one or more previous

1 Lovejoy, A. O., Reflections of a Temporalist on the New Realism. /. of Phil.,
Pnychol., etc., 1911, 8, 589-599.

368 ILLUSORY EXPERIENCE

judgments is that which carries the difficulty, and I look on a con-
flict of judgments precisely as on any other prepositional conflict.

If now it is asked how one surveying the whole situation (as
if this were possible without implicitly or subconsciously making
many judgments x) can interpret it in the spirit of realism, the
answer is simple. One interprets it precisely as Professor Love-
joy probably interprets mirrored space in connection with the space
behind the mirror. The case is parallel to our own in all respects
germane to our argument; and I have yet to hear that mirrored
space gives rise to grave difficulties about reality or about "ex-
istence merely for consciousness." It does, however, show in
exactly the same degree to which hallucinations show that "the
same portion of real space can be at once both empty and filled."
Lest the opponent have no ready interpretation of such a mystery,
I will offer one.

Mirror-space is a cross-section of ordinary space and of the bodies
therein ; it has all the pure-space properties of the other save for
a geometrically definable mode of reversal, and it includes the
surfaces and shapes of the mirrored objects (reversed) with their
colors. But it does not include the ponderability and many
other physical properties of these objects. As in the case of all
cross-sections (of which one case is consciousness), a new mani-
fold is formed, consisting of certain parts, and integral parts, of
objects; and here, as usually, such partition and analysis of the
objects reveals that they are made up of elements which are not
in the popular sense 'physical,' but rather conceptual or mathe-
matical, or, as I prefer to say, neutral — because neither physical
nor mental. These mirrored parts remain objects, but they lack
many properties/such as ponderability, et ccetera, which pertain only
to the original entire object in its natural state of organization. All
this is precisely what natural science has discovered from its own
analysis of the same objects; — matter analyzes out completely
into mathematical entities, and leaves no residue by way of little

1 Cf . the fourth chapter of Schopenhauer's Fourfold Root of the Principle of
Sufficient Reason.

REALITY 369

material brickbats. A block of wood is ponderable, et ccetera,
but the shape, volume, physical masses, and electrical charges
of which it is composed are not ponderable ; ponderability being
a property of, and deducible from, just these things in their
organized completeness. This circumstance gives rise to what
Professor Lovejoy and others have remarked as the "wide deno-
tation which the realist gives to the term object," a denotation
which physical science itself imposes. It is true that in the case
of mirrored space reproduction is involved, but it is a reproduc-
tion of identicals without change of their 'nature'; and while
it presents, indeed, interesting problems concerning particulars
and universals, it affords no footing for speculations as to a sub-
j ective realm of representations. And no more do the cases in which
consciousness involves reproduction (as it sometimes but by no
means always does); they are invariably entirely comparable to
physical reproductions. Yet they have given rise to no end of
silly talk, which, when applied to the equally eligible case of the
mirror, reads as follows : " You say that the mirror does not alter
its objects. Where then does it get its reproduction of them?
Must it not make them ? But if it makes them, they must derive
their nature from the mirror itself, and be intrinsically subjec-
tive mirror-stuff. The mirrored reality cannot be other than of
the nature of the mirroring spirit." l Or if there are two actively
imaging mirrors we read: "Data which are qualitatively dif-
ferent cannot be numerically identical; and no 'this' in the one
mirror can ever be exactly the same, qualitatively, as any 'this'
in the other mirror at the same moment, could not be unless the
two mirrors were exactly alike and occupied the same place in
space, i.e. in idealistic terms, unless every element pictured in the
two mirrors were identical ; hi which case they would be one and
the same mirror." 2 Thus it is proved to the satisfaction of idealists
that two mirrors cannot image the same object. I say, then, that
the case of hallucinations, as cited by Professor Lovejoy, is paral-

1 Cf. Carr, H. W., Proc. of the Aristotelian Soc., 1908, 8, 128.
* Cf. Drake, D., /. of Phil., Psychol, etc., 1911, 8, 369.

370 ILLUSORY EXPERIENCE

leled by such cases as that of mirrored space, wherein sundry mir-
rored objects occupy the same spatial positions as are occupied
by other 'real' objects situated behind the mirror. It is admitted
that such physical cases give rise to no difficulties about ' unreality '
or a separate subjective status. In some cases of hallucination
and illusion the error consists solely in the entertainment in the
mind of mutually contradictory propositions; and while one of
these may be for some reason preferred or assigned a superior value,
neither is more subjective than the other ; nor are both subjective,
because, as we have seen, the entire universe is brimming full of
just such mutually contradictory propositions. In other cases
of hallucination and illusion (as with the two eyes asquint) a
certain actual interpenetration of terms in relation is experienced,
which is precisely paralleled by the interpenetration of mirrored
objects with the objects behind the mirror. This is not a case of
the impossible-unthinkable realized, but it is merely a case which
shows that the neutral cross-sections of objects possess proper-
ties of interpenetrability and so forth which differ from the
properties of the entire objects. So, too, fire burns, while the idea
of fire does not, just as it still burns although the magnetic prop-
erties of flame do not. Thus, whatever practical terminology may
be found most convenient for the case of hallucinations, these
afford no grounds for an argument toward subjectivism. The real-
istic account of them is clear, simple, and straightforward, even if
it is admitted that reproduction is involved. In many if not most
cases of consciousness reproduction is not involved, and the realistic
account is still simpler.

A word may here be in place on the realist's "wide denotation
of the word object or thing." A common anti-realistic argument
is that since fire burns and the idea of fire does not, since an ob-
ject has no perspective foreshortening, while the visual image of
it has, an object has definite position in space, and the memory
image of it has not, and position in time, while the idea comes
always subsequently ; — that for these reasons realism makes
itself absurd in asserting that the idea is 'qualitatively' and in

CONSCIOUS CROSS-SECTION 371

many cases 'numerically' the object. I think that it was
probably this argument mainly which led Mr. G. E. Moore *
to the view that consciousness and its objects are distinct exis-
tents, that even color qualities are never content of sensations,
and then on to a belief hi a fundamental unresolvable relation of
awareness ; from all of which I should emphatically dissent. Now
to the argument above stated it might be replied that fire burns
but the shape of the flame does not, that an object is not fore-
shortened but that its geometrical projections are, that the object
has position in space but that, say, the nth derivative of its physi-
cal motion has not, and that while the object has position in time
the value of its physical mass is timeless. Thus the argument
rests entirely on a singularly crude brickbat notion of physical
object. This is that the object is a sort of indivisible brickbat,
of which any and every property of the object can under any and
all circumstances be predicated; or that if the brickbat has parts,
all of its properties can be predicated of each of its parts. The ab-
surdity of this is patent, and yet one can see how far such an error
may go from the astonishing words of Dr. Drake,2 who after re-
marking the many aspects presented to different spectators by a
tree, continues, "If we believe, then, that each of these 'thises'
that we have in experience is a permanent existence [he should
say a subsisting entity], we have a marvelously multiplied world.
A thousand, a million different 'thises' permanently exist as 'this
tree ! ' Disparate as they are, they cannot be squeezed down into
one simple object [indivisible brickbat], and we have a world
reduplicated ad infinitum. . . . We ['critical realists'] give him
[the 'natural man'] a simple, homogenous external world; 'nat-
ural' realism does not." I should hope not ! And I can conceive
no better advertisement for what Dr. Drake calls 'natural'
realism. But what, meanwhile, can he possibly be thinking of
the infinity of actual geometrical projections of his tree, to say
nothing of its innumerable other actual relations? Can these

1 Moore, G. E., The Refutation of Idealism. Mind, 1903, N. S., 12, 433-453.
1 Op. cit., p. 370.

372 ILLUSORY EXPERIENCE

all be "squeezed down into one simple object " ? Surely the tree
is, with its relations, all of these things : and even more. For, as
Professor Alexander says,1 of a straight stick which looks bent when
a part of it is under water, "There is illusion only if we deny that
the bent and the straight appearance in the two different sets of
conditions belong to the same stick." Certainly they belong
to it, and so does every other of its projective properties (and these
are not merely spatial, but temporal and logical as well). For
it must not be forgotten that while the object itself, if a physical
thing, is far from simple, we are always perceiving it in a compli-
cated setting of (spatial, temporal, and logical) relations, which
is a still more complicated thing. But the conscious cross-sec-
tion is always a group of the integral (neutral) components of the
object and of its innumerable relations. And since, firstly, it
is seldom possible to say just where the object itself terminates
and its relations to other entities commences (Is, for instance,
the first derivative of motion a part of the moving object or one
of the object's relations?), and since, secondly, any discrimin-
able entity may be 'object' of consciousness (i.e. a member of
the class of entities to which the nervous system specifically re-
sponds), realism has good grounds for extending the denotation of
object or thing.

This completes my argument, and more than fills the space
allotted to me in this volume. The picture which I wish to leave
is of a general universe of being in which all things physical, mental,
and logical, propositions and terms, existent and non-existent, false
and true, good and evil, real and unreal subsist. The entities of
this universe have no substance, but if the spirit is weak to under-
stand this, then let the flesh, for a season, here predicate a neutral
substance. These entities are related by external relations, and
man has as yet no just ground for doubting that the analytic
method of empirical science can proceed without limit in its in-
vestigation of this universe. The dimensions of this universe are
more than the three dimensions of space and the one of time : how
1 Alexander, S., Proc. of the Aristotelian Soc., 1910, N. S., xo, 11.

CONSCIOUS CROSS-SECTION 373

many more is not known. The line that separates the existent and
the non-existent, or the false and the true, or good and evil, or
the real from the unreal, seldom coincides, and never significantly
coincides with the line that distinguishes mental and non-mental,
subject and object, knower and known. A mind or consciousness
is a class or group of entities within the subsisting universe, as a
physical object is another class or group. One entity or complex
of entities can belong to two or more classes or groups at the same
time, as one point can be at the intersection of two or more lines :
so that an entity can be an integral part of a physical object, of a
mathematical manifold, the field of reality, and one or any number
of consciousnesses, at the same time. As the class of physical
objects is defined within the subsistent universe by principles
known to science, so the class of consciousnesses is defined within
this universe by principles which are partly known, and which
are coming to be more fully known, by empirical psychology.
A consciousness is the group of (neutral) entities to which a
nervous system, both at one moment and in the course of its life
history, responds with a specific response.

SOME REALISTIC IMPLICATIONS OF BIOLOGY

SOME REALISTIC IMPLICATIONS OF BIOLOGY
BY WALTER B. PITKIN

THE following investigation falls into two parts. The first
part is a formal analysis of several fundamental types of organic
action. Here, certain facts appear which decisively refute several
biological, psychological, and metaphysical theories. The second
part, which begins with Section IV, will offer an incomplete formal
analysis of the cognitive situation and a hypothesis of consciousness.

The five chief points herein to be urged are : (1) That theories
about life and mind have generally been based upon erroneous,
more or less naiive opinions about the nature of things and relations.
(2) That anti-realistic hypotheses in biology or derived from biol-
ogy owe their strength either to (a) the erroneous opinions mentioned
above or (6) a false emphasis upon the facts of introspective psy-
chology. (3) That the ordinary methods of analysis 'and re-
search pursued by natural scientists are rapidly yielding a coherent,
adequate account of life and mind, without invoking either intui-
tion or transcendent entities. (4) That a formal analysis of the
organic situation discloses three basic factors (each complex) re-
lated as are the three factors in projective geometry. (5) That
an interpretation of the cognitive function in terms of these three
factors largely reconciles at least four modern theories of conscious-
ness and also makes clear how errors, illusions, and hallucinations
are not made by consciousness nor are peculiar to it, but are
necessary features of a projected physical system.

This last topic is introduced, less to convince the reader than
to offer a definite, positive theory which may be criticized. Real-
istic studies have, by necessity, been overwhelmingly polemical ;
but they should not continue so. Reconstruction must begin ; and
a theory of life and mind must be worked out which dispenses with
the old, discredited categories of idealistic psychology, such as

377

378 IMPLICATIONS OF BIOLOGY

'mental states/ 'subject-object polarity/ 'creative synthesis/ and
the like. Now it is evident that the first attempts to slough off
these notions will be not only difficult, but full of strange writhings.
They will be no less violent than an endeavor to exchange the parts
of speech in one's native tongue and to use nouns for adverbs, or
adverbs for prepositions. For the older manner of thinking is
woven into our unconscious 'universes of discourse.' Because of
this, any genuinely realistic hypothesis of consciousness to-day
must be obscure; and it is almost certain to contain difficulties
which the author himself cannot clearly sense. It must, there-
fore, be submitted in fear and trembling.

THE BIOLOGICAL ATTACK ON REALISM

THE strongest influences against realism to-day emanate from
the biological sciences. Only a few years ago, it was physics and
mathematics which made the natural world-view seem unten-
able ; and before them it was logic and psychology. The objec-
tions which these sciences have brought have been cleared away
in the previous essays of this volume; the inaccurate descrip-
tions and defective logic underlying those objections must now
be moderately clear. But, from a new quarter, there arises a
host of adversaries, declaring that the unanswerable disproof of
realism is found in the simple life-processes.

At least three lines of research lead up to this assertion. The
first lies wholly within the biological sciences ; it has been followed
by men whose intimate knowledge of life-processes commands re-
spect. Hans Driesch is the modern leader of this movement.
His Gifford lectures on 'The Science and Philosophy of the Or-
ganism' 1 are a remarkably thorough, systematic attempt to
establish idealistic vitalism on biological evidences.

The second line of research proceeds from psychology to biology.
The immediate data of consciousness, when properly described,

1 2 vols., London, 1908.

THE BIOLOGICAL ATTACK 379

afford a new basis for interpreting life processes ; and the latter,
when reinterpreted, reveal a cosmos not composed of distinct
characters, a flux wherein everything interpenetrates everything
else. All distinctions are products of a 'vital force' and serve
only for organic controls. To this hypothesis comes Henri Berg-
son, in his now famous trilogy.1 The third line of research yields
a formal analysis of the fundamental biological situation. It is
not a theory about life-processes. It is rather a simple description
of (a) the situation in which conduct arises, and (6) the direction
and character of conduct. It bears the same relation to the bio-
logical explanation of conduct that the equation and the graph
of a given trajectory bear to an explanation of the chemical con-
stitution of the nitroglycerine that hurls the projectile. The
leader of this movement is Dewey, whose book on 'The Study of
Ethics ' 2 was the first formal analysis 3 of conduct ever attempted.
These three lines of research terminate in as many anti-realistic
hypotheses. Driesch concludes that the entire 'content' of ex-
perience is created by the Ego, in the same manner as Kant held.
Bergson believes that there is an objective flux which constitutes
the environment of the 'vital force/ and also that in primitive
perception some of its characters are immediately presented; but
he insists that all discreteness is produced by the 'vital force/
and hence that not only the differences between individual per-
cepts, but also all concepts, are 'static/ 'condensations/ 'con-
venient abstractions.' Dewey is by all odds the most realistic.
He accepts as an ultimate the biological situation (i.e. an environ-
ment and an 'agent/ the former acting upon the latter, and the
latter reacting upon the former) ; moreover, he regards perception
as a 'natural event' on a par with a thunderstorm, in that its
characters are not created by the percipient. He draws the real-
istic line, though, at cases of knowledge, which, he says, are wholly

1 Essai sur les Donnees Imm6diatea de la Conscience, Paris, 1888, Matiere et
Memoire, 1908, and L'Evolution Cr6atrice, Paris, 1909.

* Ann Arbor, 1897.

* For a fuller description of formal analysis, cf. Spaulding, 156. The meaning
of ' formal ' in this connection must be carefully guarded.

380 IMPLICATIONS OF BIOLOGY

different from perception. Theories and their constituent ideas
are genuine constructions of the thinker. They have no existence
apart from him, any more than a mowing machine has. The
stuff out of which the machine is fashioned is indubitably physical,
but the arrangement and all its efficiencies are a result of thinking ;
it would therefore be absurd to allege that somehow in the stuff
the mowing machine is contained (implied, potentially present).
Now, what iron and hickory are to the machine, percepts are to
knowledge.

The realist cannot count his case won until he has disproved
the anti-realistic inf erences of these three investigators. He must
therefore scrutinize just enough of their deductions to show the
complete independence of all things thought of.1 Now, to discuss
the relation of agent to environment, we must agree on a descrip-
tion of the situation under debate. This description should enable
us to effect a division of the question, if the latter proves complex.
If the environment is a hazy entity, it must be cleared up ; if it
is a complex, its simple constituents must be discerned and their
relations to one another indicated. Likewise with the agent :
whether we are to mean by that term the whole organism, or a part
of it, or something 'behind' the organism, or something else must
be settled before another step can be profitably taken. All this,
of course, is no explanation of any biological happening. It is
only a clear visioning of what happens. It is not complete insight,
but only accurate description of as much as is described at all. To
give this is our first task.

II

FORMAL ANALYSIS OF THE BIOLOGICAL SITUATION

NATURALLY inspecting animals and their circumstances of life,
we find : —

1 : that they exist in a world larger than themselves.
2: that this world sets for them certain difficulties.

1 For the sense in which ' independence ' is here used, cf. Perry's essay.

THE SITUATION ANALYZED 381

3 : that some individuals overcome these difficulties.

4 : that those which signally overcome the difficulties differ, in
some observable respects, from those which do not.

5 : that the two phases of an organism, namely its structure and
its functions, are very closely correlated, and perhaps absolutely.
(Structure is the arrangement and character of the parts or ele-
ments of the organism ; function is the natural operation of these
under the circumstances of existence.)

6: that both structure and function are, in all save the very
lowest organisms, differentiated. There are, in a given animal,
many structures, and each does something peculiar. Each such
structure is called an organ.

7: that what a given organ does depends upon at least three
factors : (a) its own total character (especially its acquired char-
acter) ; (&) its relation to other organs (i.e. its connections with
them and their own total character at the given moment; and
(c) those influences from the environment which affect the organ,
at the moment.

8 : that both the structure and the function of an organ vary,
in a more or less measurable way, with some variations in the ex-
ternal stimulus.

All these are simply observed features of the situation ; they are
not an explanation of anything, nor does our acceptance of them
depend upon any particular theory of life processes. On the con-
trary, all theories about the latter are only attempts to explain
these peculiarities in their many concrete forms. A hypothesis
which asserted or implied the non-existence of any one of these
would not be dealing with the biological situation at all. This
becomes more obvious, as soon as we force a special interpretation.
Suppose, for instance, that a biologist were asked to explain how
an elephant comes to grow a trunk. If the learned man assured
you that there wasn't any elephant, or that what you called its
trunk was only a product of your own mind, you might be inter-
ested and perhaps enlightened ; but you would not be enlightened

382 IMPLICATIONS OF BIOLOGY

on the subject of your inquiry. His assertion would be irrele-
vant Were it true, it would not dispense with your question;
for you would still wish to know how in a world of illusion
an illusory elephant comes to have an illusory trunk. The crea-
ture surely does grow a trunk, no less regularly than before ; and
the trunk has certain definite, constant peculiarities. You ask
for an explanation of this particular situation. It is unprofit-
able, then, to be told that the terms of this dynamic relation have
a different status from that which you may have assigned to
them ; for you have asked, not about the status of the terms, but
about the origins of the relation between them.

Let us first consider the reciprocal influences of the organism
and the environment. In what manner does the one alter the
other? Precisely what is changed, in a given situation? Wo
must describe specific reactions and their stimuli, and find out
whether they reciprocally transform each other; and, if so, what
the general nature of the transformation is : and, secondly, we
must ask whether the transformations we may find are sufficiently
profound to warrant the inference that any (or certain) entities
which men perceive or establish by thinking are strictly consti-
tuted by the act of perceiving or thinking.

1. Simple description of some stimuli and the reactions they in-
duce . —

A. What is the descriptive distinction between stimulus and reac-
tion ?

In distinguishing stimulus from reaction, we have no right to
smuggle in any characteristic which is only inferred. This rule
of formal analysis compels us to admit that the precise limit of a
stimulus action cannot be pointed out ; nor can the origin of the
reaction. We are not even free to say that all processes within
the organism are reaction processes, in the strict sense. Many
may be merely conduction processes, serving only to carry a stimu-
lus to some further sensitive region, at which the reaction is set
up. A number of organic events seem to be this type. For in-

STIMULUS AND REACTION 383

stance, some experiments indicate that the nucleus of the single
cell is the sensitive part from which all reflexes originate. If this
is true, then the other regions of the cell serve as mechanisms for
bringing in to the nucleus certain influences from without, or else
for effecting the adjustment which the nucleus initiates. But, as
Pfeffer and Delage have pointed out, this is only a hypothesis;
and it has some serious difficulties. We cannot admit it here, then.

Nevertheless we can describe stimulus and reaction accurately
enough to distinguish them without equivocation. The stimulus
includes whatever acts upon the reagent so that the latter acts
with reference to the former. That is, not every influence acting
upon the organism is a stimulus. Thus, the pull of the moon upon
a man's heart and the action of terrestrial magnetism on his neural
currents would not be stimuli. For neither sets up any reaction,
nor does the organism behave in any specific manner toward either.
But the action of gravitation on the semicircular canals is a stimu-
lus, and so too in geotropic plants. For here there is a reaction
to differences in gravitational pulls.

Another point to notice is that stimuli are not exclusively extra-
organic, nor are they nothing but stimuli. The poisons generated
during fatigue are, with reference to the moments succeeding their
generation, true stimuli. So far as we are now concerned, they
do not differ generically from bacterial toxins. There may be
instances of a given influence being, at the same instant, both a
stimulus and the product of a reaction. In such a situation there
is no mystery or contradiction ; for the influence is not both stimu-
lus and reaction product in the same sense. It is the stimulus of
one reaction and the product of another, the two reactions going
on simultaneously.

The reaction is whatever alters some part, element, or action
of the organism in relation to the stimulus. It is not to be identi-
fied with every effect of an extra-organic influence upon the body.
Many a bodily motion and change is not a reaction, although its
cause is an incoming force. For instance : a man falls to the
ground under the shock of a violent blow in the face ; his motion

384 IMPLICATIONS OF BIOLOGY

is no reaction, for it is a mere effect, not that peculiar kind which
refers back to its cause somehow. Were he to struggle to keep
his footing, though, he would in that measure be reacting.

There is nothing to hinder a given reaction from being, at one
and the same time, an adjustment to a stimulus and also a stimulus
to other reagents in the same organism. For instance, reactions of
the salivary glands at the sight of food certainly have for their
stimulus the visual reaction ; and, in turn, they become the stimu-
lus of various gastric glands. This does not indicate that stimu-
lus and reaction are 'purely relative' terms, in the sense that the
distinction between them is not genuine, and that a stimulus is
not 'really' a stimulus, but now one thing and now another, ac-
cording to the way one happens to look at it. Like every other
real entity, stimulus and reaction stand individually in innu-
merable relations at every instant, without any loss of their dis-
tinctness. So they appear, and so must they be described.

About the reaction, we may say more. Its peculiar character
appears best in its outcome for the organism; and this outcome
is often conspicuous and may be described in terms of the stimulus.
Thus, the lungs select oxygen and pass it on to the blood ; the heart
drives on the blood; the eye regulates certain motor reflexes;
and so on. The special task, in each case, is most sharply dis-
tinguished, not by the details of the reaction process but by its final
relation to the stimulus which has set up the reaction.

In looking to this final relation between reaction and stimulus,
we are not implying that this relation has been 'aimed at.' The
final relation is not necessarily a finalistic one. All that we assume
is the rather obvious fact that the character of any process cannot be
clearly read off until the whole process is before us. Were we de-
scribing a dynamite explosion, we should have to look to the final
phase of it, in order to know what had happened, and what a dyna-
mite explosion is. Were we to halt in our description at the very
instant when the unstable molecules of the stuff were disintegrated,
we should not be describing the explosion. Everything deriving
peculiarly from that disintegration must be inspected and specified ;

TYPES OF REACTION 385

e.g. the kind, volume, temperature, pressure, incandescence, etc.,
of the gases generated.

B. Some Types of Reaction. — This final relation can, in many
cases, be very precisely indicated, although the means of attaining
it are absolutely unknown. It has many distinct types, seven of
which are the following :

(a) Adjustment ; (i~) partial and (ii) total.

(6) Selection

(c) Conduction

(d) Transformation; i. Non-Constitutive ii. Constitutive

a. Selective
/?. Additive

(e) Resistance

(f) Reception
(gf) Retention

These last two will not be discussed at present ; they involve
psychological issues, all of which are to be avoided until the prob-
lem of consciousness is broached, in the closing section of this essay.
The other five types of reaction may be briefly described as follows :

(a) The outcome of some reactions is to put (i) one part of the
organism into a new relation with some part of the stimulus ; and
of others to put (ii) the organism, as a whole, into a new relation
with some part of the stimulus.

(t) The lens of the eye reacts so as to throw a sharp image of the
stimulus upon the retina. Within this act the organism as a whole
does not alter its relation to the environment. And only one fea-
ture of the stimulus is reacted to. The color, intensity, and spe-
cific character are inefficient with respect to the ciliary reflex.

(ii) The leg muscles react so as to bring the entire animal to-
ward or away from (i) the origin of the stimulus or (ii) the im-
mediate influence of the stimulus.

(6) The lungs select oxygen from the inhaled air and reject vari-
ous other gases. The ear picks out air vibrations and is indifferent
to ether vibrations.

(c) The afferent nerves carry certain characters of peripheral
2o

386 IMPLICATIONS OF BIOLOGY

conditions to the central nervous system. Whether these periph-
eral conditions are identical with characters of the stimuli will
be later discussed.

(d) The stomach adds solvents to the material taken in, and the
material is thereby altered no less qualitatively than hydrogen is
altered when it is combined with oxygen. The alteration is not
'absolute/ for the constituents of the new mixture can all be
separated and identified. It does, however, involve genuine now
qualities.

This fact raises at once a question of terminology. Shall we
designate as transformative every reaction which develops a qual-
ity (character) which was not present and efficient in the same
sense in the complex stimulus? A question of terminology, in-
deed ! But, like many others, one on which many vital distinc-
tions of philosophy and science turn. It is largely the unclarity
of words here that has brought about differences of opinion about
the relation we are studying.

Now, for present purposes, it is enough to demarcate three con-
ceivable types of relation between stimulus and reaction : —
Species A : Non-constitutive.

Type 1 ; The reaction simply releases one part or element or

phase of the stimulus.
Type 2; The reaction adds some part or element or phase of

something else to the stimulus.
Species B : Constitutive.

Reaction and stimulus are simply two phases of one unitary
process, as the two poles of a magnet are. Just as the posi-
tive pole does not take something from nor add something to
the negative pole, so the reaction does not consist in selecting
something from the stimulus or adding something to it. Each
gets its character from its relation to the other.
These three types represent distinctions which have been drawn.
So long as they are going to be employed by anybody, their dif-
ferences should be recognized and appropriately named.
Now, it strikes me that Type Al should not be called a qualita-

TYPES OF REACTION 387

tive transformation. To label it thus is to blur over several highly
important differences. There is no denying that when an element
is released from a complex, both the element and the remaining
complex behave as they did not behave before. There is a change
from behavior of one character, if you will, to behavior of another
character ; but this is not at all identical with a qualitative change
of the behaving entity. The change of behavior may be one and
the same with a change of the entity's position in a complex.

For instance, a dog is frequently docile when at large and
vicious only when tethered. Would anybody say, though, that in
turning the dog loose, we change his quality ? Hardly : for the
releasing, as such, does not constitute the difference in the animal's
quality of action. The animal changes its conduct, under different
circumstances; but the act of changing the circumstances is not
identical with the dog's changing its behavior. Neither is the
animal's doing different things under different circumstances a
sign of a change of nature. To act now thus and now so under
identical conditions, though, would indicate such a change. The
new circumstances (after the untethering) bring the dog into new
relations ; and it is in response to these latter that it now behaves
differently. The distinction is fine, to be sure, but none the less
genuine. It appears more clearly in biological cases.

If a reaction selects a part of a stimulus complex, this means
simply that it picks the part out, just as one might take the dog
out of the barn. But, having so removed it from its setting, it
has perforce let it into some other settings. The lungs not only
pick out oxygen, but, having done so, turn it loose in the blood.
And now the oxygen, released from previous restraints, as it were,
does things which it previously could not do. But did the selec-
tive reaction endow it with these newly developed efficiencies?
Not at all, and we can prove empirically that it did not, simply by
separating the oxygen in the laboratory and discovering all those
very qualities. It is, therefore, the positional relation of the part
to various complexes that determines the results, quite independ-
ently of the way in which that positional relation has been brought

388 IMPLICATIONS OF BIOLOGY

about. Of a truly selective reaction, then, we must say that it
does not bestow qualities upon entities (stimuli) or take them away.

Type A2 may reasonably be called a 'qualitative transforma-
tion' in one restricted and somewhat questionable sense. The dog
which is savage when tied is not changed qualitatively by the mere
act of being released, but only by the new situation into which it
enters when free. If, however, the dog is overfed, while tethered,
it may grow less ugly. Now, that which is added changes the
quality of the dog's conduct, quite independently of any change
in the dog's relation to the barn or the tether or the persons barked
at. The important point to note here is that the addition is made,
not to the total complex (dog-on-rope-in-barn), but to the canine
part of it alone; and, secondly, that adding does more than bring
the part (dog) into a new relation with other parts of that or other
complexes; it involves an internal change of the particular part
involved. The position of the dog, as a whole, to its rope, the barn,
and the passers-by is not at all changed by overfeeding; hence
so long as we regard the dog simply as a part of such a complex, it is a
logical simple in that particular complex ; and the change wrought
is therefore qualitative with respect to that particular complex. The
change does not consist in a change of its positional or other rela-
tions to the complex. So, too, the stomach genuinely qualifies
the food it receives, inasmuch as it lends it chemicals and efficien-
cies which were not there previously, either in any element as such
or in the complex as such, in any sense.

Obviously, though, such a transformation is 'qualitative' only
because we treat as a simple a factor which we have not analyzed.
In the complex, dog-on-rope-in-barn, the dog is an element. Now
an element is not identical with a simple, in the sense in which we
speak of the dog here as an element. But, for many purposes, we
do not need to analyze the dog, in so far as we are dealing merely
with this particular complex. When this is the case, we reckon
the dog as simple, within the limits of our inquiry, and hence we
must say that every change in the animal is a change in a simple,
and hence an 'internal,' 'qualitative' change. So long as we un-

TYPES OF REACTION 889

derstand clearly the restricted universe of discourse within which
this proposition holds, there is no harm in asserting it.

Type B is indisputably a qualitative transformation, and more.
The reaction and the stimulus are not only modified ; they are in
some degree constituted by it. Their nature is just this being a
pole of a polar complex. Thus it is radically unlike the suppressed
qualities of Al or the added qualities and the pseudo-simple changes
of A2.

I shall hereafter speak of transformations as of Type Al, A2,
and B. Type Al is identical with the simple selective reaction,
listed as No. 2 in the enumeration of certain fundamental reaction
types. For the sake of a historical understanding of so-called
'transformations,' I shall tolerate this cross-division.

(e) The last type of reaction here to be described is resistance.
The organism, affected by some injurious stimulus, internal or
external, does not avoid it, as in ordinary adjustment ; it builds
up structures which intervene between the stimulus and the organic
parts which the stimulus injures. Frequently this process appears
to be incidental to the regeneration process of an injured structure.
(In this manner Ehrlich explains antitoxins.) In other cases this
is not evident ; as in the fox's growing of thicker fur in winter, in
the pigmentation of skin under extremely bright sunlight, etc.
And it is quite obscure in the case of attention to cognized entities,
where some kind of resistance is offered to all other elements tend-
ing toward the cognitive field.

Let us now consider each of these five reaction types in more
detail.

a. Simple Adjustment. — Stahl has demonstrated * that the direc-
tion of incident light determines the position of the first cell wall
in the developing spore of Equisetum. And Pfeffer's experiments
show that the same factor, direction of light, is what fixes the plane
of symmetry in the growth of Marchantia.2 The roots of plants

1 Berichte d. Bot. Ges., 1885, 334.

2 Sachs's Arbeiten, i. 92.

390 IMPLICATIONS OF BIOLOGY

regularly turn in the direction of the gravitational pull, and the
stems away from it. Some species are so sensitive that a very
slight change of position sets up a reaction which readjusts the
growths to their previous directions. Driesch's experiments with
Sertularia also show this very prettily.

Now, let us try to explain absolutely nothing, but only describe
what has happened. We note the following three points :

(i) It is not the color, nor the intensity, nor the duration of the
stimulus which causes the adjustment of organic parts into definite
positions. It is only the direction of the stimulus. If, then, we
are to speak of the color or the intensity of light as a cause of other
types of reactions, we must now speak of direction as being no less a
true cause in the present case. It is, in short, not a metaphysical
hypothesis to say that direction is a cause ; it is pure formal analy-
sis. Direction is the necessary antecedent concomitant variant.
Descriptively, it is an efficiency, no less than the mass or the elec-
tric charge is.

This fact is not altered by the other alleged fact, that, hi adjust-
ments to the direction of gravity, this effect is brought about by
the sinking of the heavier cells within the organism and the rising
of the lighter kinds, which latter are central in the roots and pe-
ripheral in the stems.1 Suppose this does occur ; is it not the very
same happening which we have just been describing, in another
manner ? It is the direction of the gravity pull which causes the
direction of the motion of the cells. And the proof is as before :
vary the direction of the stimulus, and you vary correspondingly
the directions of the moving cells.

The implication of this reaches far beyond biology. It touches
physics and mechanics. Here, too, the efficiency of a force cannot

1 1 cite this hypothesis only because it presents the strongest case against the
view above set forth. But it is strong only in form ; to-day it cannot cover the facts
of geotropism. Pfeffer and Czapek have proved that only those cells within 1 mm.
of the tip of a root are sensitive to geotropic influences. The position of this tip
determines that of the whole root, and its removal suspends the geotropism until
its regeneration is completed. In the light of this fact, we cannot admit that the
mere pressure of cell against cell accomplishes the reaction.

SIMPLE ADJUSTMENT 391

be limited to the pure quantity of energy involved ; for, if it were,
then no regular relation would appear between the direction of a
resultant force and the directions of its components. Vector analy-
sis, however, reveals a simple and inflexible relation there. It is
the direction, and not the velocity, mass, acceleration, or retarda-
tion exclusively which determines the path of a body acted upon
by two or more forces in a pure mechanical system; and, in a
physical system, this same direction value causes the trans-
formation of kinetic into potential. That is, 'energy is consumed'
in changing the direction of a body ; or, otherwise stated, a force
in one direction is efficiently equal to (reduces to) a smaller force
after re-direction. And here, even as in the case of the developing
spores, concomitant variation proves that it is the direction which
is at least partly the cause of this. And by 'cause' we mean
precisely this sort of necessary prior concomitant of a given
event.

(ii) The reaction is generically like the stimulus. That is, the
latter is a direction (or, more accurately, a force whose directive
character is the determining one in the particular instance), and
so too is the former. This is a most important feature of the situ-
ation, for it sets a very narrow limit to our interpretation of the
relation between stimulus and reaction, in this case. We can say
only that the specific stimulus causes a re-direction of motions in
the reagent ; it does not give rise to any new qualities, in the proper
sense of the term ' quality.' There is a genuine change, of course,
but it is from one direction to another ; not from one quality to
another.

We are now ready to state at least one restriction which must be
put upon all three varieties of biological anti-realism. They can-
not maintain that every reaction involves a qualitative transfor-
mation of its stimulus, unless they define 'quality' so as to make it
absolutely everything that makes any kind of a difference. Such
a definition reduces 'quality' and 'difference' to almost inter-
changeable terms. It does violence to good usage and offers no
new philosophical convenience, by way of recompense. How its

392 IMPLICATIONS OF BIOLOGY

champions defend it is hard to see. But if they do employ it,
they must still admit that the reaction does not necessarily involve
a generic change of quality, but only a change from one individual
quality to another of the same genus.

(Hi) The differences of reaction are related to one another in the
same way as are the differences of stimulus. For instance, if a
specimen of Sertularia be tipped ten degrees eastward from per-
pendicular, the difference between gravitational stimulus A (the
one before the tipping) and stimulus B (the new one) is just that
ten-degree angle. And this same angle constitutes the difference
between the reaction to stimulus A and the reaction to stimulus
B. Obvious as this is, its import has, so far as I know, never been
sensed by those biological philosophers who are urging that the
'quality' of the reaction is other than that of the stimulus, or
that the reaction qualitatively transforms the stimulus. There
is not simply a one-to-one correspondence between stimuli and
adjustment reactions (two qualitatively heterogeneous systems);
there is also a qualitative identity between the differences urithin
each system. The difference between stimulus A and stimulus B
is not simply paralleled by a difference between reaction A and
reaction B ; but the former difference is identical with the latter,
in kind and in measure. The implications of this fact appear,
as soon as one seeks to interpret the two systems as heterogeneous,
as we shall soon allow the anti-realist to do.1

Let us consider a case of adjustment in human life. One of the
least obscure is the ciliary reflex. What happens here ? The lens
of the eye is thickened or flattened, so that an image with sharp
outlines is thrown upon the retina. Now, however this adjust-
ment is effected, it is determined, not by the color nor the bright-
ness nor by the shape nor by the size nor by the mere grouping of
elements in the visual field, but by the distance of the element at-
tended to from the lens (or by some function of that distance).
This is none the less true, even though the adjustment is accom-
plished by trial and error methods ; i.e. by running rapidly through

i Cf., 428 ff.

SIMPLE SELECTION 393

the whole range of adjustments until that one is hit upon which
gives a sharp retinal image. Suppose this is what happens. Then
the adjustment is determined by the distance between the edges
of overlapping images on the retina; and these distances, at
each individual adjustment of the lens, are functions of the dis-
tance of the external object, as can be proved by simple experi-
ments in optics. The extent of the ciliary contraction or expan-
sion, therefore, is a function of a function of the object's distance ;
in other words, a second derivative. As in the previous cases, so
here; distance is a genuine efficiency, and the reaction to it is
generally like it, namely an adjustment of distances (as well as an
adjustment to distances). And, finally, the differences between
the stimuli are of the same kind as the differences between the
reactions; both are distance-complexes.

&. Simple Selection. — This exhibits two phases : first, the dis-
integration of an objective complex (the stimulus), so that the
element to be selected is released from the complex ; and, secondly,
the rejection of those other elements of the complex which are not
chosen. The former activity sometimes involves the addition of
something to the stimulant complex, as a means to the disintegra-
tion of the latter ; and this addition may be, in one case, to the
element that is to be selected, and, in another case, to the total
complex. In all instances, however, the outcome is the same gen-
erically ; one element is selected and the others discarded.

Simple selection is so familiar that a detailed description of
cases would be needless. It will suffice to name a few striking
instances. The eye selects light waves from the total medium act-
ing upon it ; the ear selects from that same medium only air waves.
The lungs select oxygen, the digestive tract carbohydrates, and
so on. Now this relation is accurately stated only by just this
same term, 'selection.' The reaction is the picking out of a part
of that which sets up the reaction. It would be foolish to ask
whether the reaction resembles the stimulus here ; the reaction is
the process of selecting. And such a process no more resembles

394 IMPLICATIONS OF BIOLOGY

that from which something is selected than the motions of a wo-
man's arm in sorting strawberries resemble the fruit. What she
picks out, though, not only resembles part of the mass she has
been picking over ; it actually is a part of that mass, transferred
to a new situation and relieved of sundry undesirable elements, to
wit, the stems and the soft spots.

No matter how closely one were to inspect the selecting process,
then, one would never discover in it anything like that which it is
selecting. But were one to look at the material with which the
process operates, one would certainly discover, at every step in
the operation, some element of it in some arrangement. These
elements, however, might be so disposed that, at certain instants
in the process, their identification would be impossible. This
difficulty, though, is not even the flimsiest evidence against the
real presence of the material in the process. The real test of its
presence is indescribably simple ; it is this : if you can put it into
the process and later take it out, you know it has been there, in one
mode or another, during the process.

In the grosser reactions this test is easily made. A man is put
in a room and oxygen released; in due season his blood will be
carrying more oxygen. The oxygen is taken away and his air
supply cut off ; soon the quantity of oxygen in his blood drops to
the danger point. So too with the chemicals he selects from food ;
they reappear, after digestion, in the blood, and their entire prog-
ress may be accurately traced from the table to the living tissue.
Can the same be proved in any of the finer, more elusive processes
of the reacting organism? Does it occur, for example, in the
sensing of colors, sounds, odors, etc. ? Or are all these reactions
of a wholly different order? This question will be faced later.
Just now, I wish only to point out that there is a profound differ-
ence between selection and adjustment, a difference great enough
to throw grave doubt upon any hypothesis which deals with re-
actions in general. In adjustment, the reaction is most often generi-
cally one with the stimulus; in selection, it bears not the remotest
resemblance to it. In the picking up of oxygen, there is nothing

SIMPLE SELECTION 395

resembling oxygen ; but in the ciliary reflex and like cases, the reac-
tion consists hi a complex of positions and directions, and so too
does the stimulus. In adjustment, again, no element of the stimu-
lant complex is identically taken up ; and in selection, on the other
hand, the operations are not the operations of oxygen, in whole or
in part. In adjustment, there is no transforming of the stimulus,
but only the taking of an attitude or posture toward it.

Plainly, adjustment of the sort we have been describing is not
what biological philosophers refer to when they call life itself an
adjustment to an environment. In their sense of the term, ad-
justment is exceedingly vague and embraces all manners of ma-
nipulations, changes, and remodeling of external entities. Is it
not extremely probable that, through this loose usage, the notion
has been strengthened that adjustments are intrinsically trans-
formative processes ?

It must be evident, by this time, that we have here to do, not only
with two heterogeneous types of reaction, but also with two very
different orders of stimuli. In selection, the stimulus is that
which enters the organism; in adjustment, it is commonly an
order outside of the organism which, in the ordinary sense of the
verb, does not enter the organism at all, but merely acts char-
acteristically upon it. This difference is not wiped out by the
fact that often the organism selects stimuli for adjustment, selects
from among its own various possible adjustments, and adjusts
for a certain selection. Such compound processes do not rob their
constituent phases of their proper characteristics, any more than
water abolishes oxygen and hydrogen.

There is a further peculiarity of the relation between stimulus
and reaction, hi the case of adjustments. The adjusting reaction
is not directed to that effect of the stimulus which enters the or-
ganism, but rather to the stimulus 'itself (in common par-
lance, to the 'external object' or quality). Thus, although the
ciliary reflex may be regulated by the distances between the edges
of overlapping images on the retina, the direction and distance of
the motions of the muscle fibers in the ciliary muscle regulate the

396 IMPLICATIONS OF BIOLOGY

directions and distances of the causes of those overlapping images.
Please catch this distinction. It is of tremendous importance,
and is not theory, but simple description. These causes are, at
the instant of adjustment, not yet a genuine stimulus; they will
be, in the very next moment. And the lens is altered, so as to alter
the causes of the images, viz. the as yet unsensed ether waves
beyond the lens. The alteration aimed at does not take place
outside of the organism ; but it does occur prior to the stimulation
by that which is so altered. That is, the light waves are changed in
the lens, not directly on the retina where they later produce images.
This operation is analagous to that of a stomach which might re-
model food before the latter reached it, developing a special organ
for this task. It may be, of course, that teeth and salivary glands
are the results of such an adjustment ; certainly their relation to
the digestive process is much like that of the ciliary reflex to the
visual process. But such speculation carries us beyond formal
analysis. All that we can say, in pure description, is that adjust-
ment reactions, unlike the purely selective, are directed backward
from the immediate stimuli to the causes of these. Or, to state
it in general form, adjustment reactions (and a fortiori the struc-
tural development of their mechanism) deal with the cause-effect
relation.1

c. Conduction. — This relation between stimulus and reaction
was doubtful until recently, and for the very good reason that it
has not been directly observable, as adjustments and selections
and transformations have been. With the improvement of re-
cording instruments, though, it has rapidly been clearing up ; and
to-day, as Holt has pointed out,2 we are in a position to assert that,
in some cases at least, the reaction consists in the transmission

1 To accept this description, we do not have to know what the cause-effect relation
is. It is adequately marked off by the statement that the effect is a function of
any prior variable in a real system. This does not limit causes and effects to
purely mechanical relations. Any pan* of characters, A, B, constitutes a cause-
effect pair if A is at (in) t, and B is at (in) if, and B = (/) A.

» Holt, 321 ff.

CONDUCTION 397

of the periodicity of the stimulus. Lord Rayleigh's experiments,
cited by Holt, prove that sounds of lower pitch than 128 d.v. a
sec. — and possibly some of higher — cannot be localized by means
of differences between their heard intensities, but can on the basis
of their relative phases, as they enter the ears. This means, of
course, that this difference of phase is sensed.1 Sherrington has
shown that nerve impulses follow the rhythm of stimulation, up
to at least five hundred shocks per second; and there are good
reasons for believing that, with better means of registering, this
symmetry would be found to hold at higher rates.

I believe, however, that we need not wait on such technical im-
provements in order to describe the possibilities of conduction.
In the adaptation of the flatfish to its background, we have the most
striking of a host of instances which show directly the relation of
approximate identity between the stimulus and the conductive
reaction. The behavior of the flatfish is so remarkable that I shall
describe it at length. As it beautifully illustrates selection and
spatial transformation, I shall mar the logical order of my topics
by here describing these other processes also, which are phases of
one triplex reaction in this curious animal.

There has recently appeared a monograph on this subject 2 by
Francis B. Sumner, of the United States Fisheries Laboratory at
Woods Hole, Massachusetts. It interests me peculiarly because
it confirms, by experimentation, my previously published hypothe-
sis about the retinal image and the imitative reflex.3 My own in-
ferences were reached by an analysis of some obscure psychological

1 Taken in isolation, this fact does not indicate homogeneous propagation of the
stimulus, by any method of formal analysis ; for the sensing of phase differences
might, so far as the direct observation goes, occur in the peripheral organs. Ray-
leigh's statement is therefore an inference, not pure description. It 'is, however,
a very sound inference. For there are numerous other experiments, especially
with motor nerves, in which just such symmetrical conduction has been observed,
and not inferred. I allude to these above and again refer the reader to Holt.

* The Adjustment of Flatfishes to Various Backgrounds, Journal of Experimental
Zoology, 10, No. 4.

*/. of Phil., Psychol., etc., 1, 92, 204. The second of these papers was read
in part, before the American Philosophical Association, December, 1909.

398 IMPLICATIONS OF BIOLOGY

events. In order to follow the analysis, the reader had to observe
those events with severe accuracy; but this seems to have been
not at all easy. In Simmer's data, though, no such difficulty arises.

The flatfish changes its hue to conform to the color of the back-
grounds on which the creature happens to lie. For a tune biolo-
gists supposed that this adaptation was effected by some direct
photochemism upon the skin. But, a quarter century ago, Pou-
chet proved that it was brought about through the functioning of
the eye. He found that blinded fish do not change their color adap-
tively. Interesting and significant as this discovery is, however,
it does not tell us anything about the most astounding behavior
of the flatfish, namely, its adaptation of its own geometrical skin-
patterns to copy the geometrical pattern of the sea-bottom upon
which it rests. By one of those freaks of circumstance which are
only too common in all scientific fields, nobody paid serious atten-
tion to this phenomenon until Sumner approached it last year, first
at the Naples aquarium and later at Woods Hole. "In observing
a turbot," says Sumner, "I was impressed by the detailed resem-
blance which obtained between the markings of the skin and the
appearance of the gravel on which the fish rested. . . . The query
at once suggested itself : Is it a mere coincidence, or does the fish
have the power of controlling the color pattern as well as the gen-
eral color tone of the body?"

To answer this, Sumner prepared a number of backgrounds,
some reproducing various types of natural sea-bottom (fine sand,
coarse sand, fine gravel, coarse gravel, of various colors), and some
being highly unnatural geometrical patterns (checkerboard, polka
dot, stripes, screen, etc.). Placed in a tank having one of these
patterns on its bottom, the flatfish began to copy the pattern on
its back. ('Copying' does not imply 'consciousness' or 'effort,'
so far as the mere use of the word here is concerned.) The time
required to complete the imitation varied.

This time ranged from a few seconds to several days. A change
involving the almost complete withdrawal from view of the skin
pigments in a dark specimen probably required the longest period.

THE FLATFISH 399

In general, . . . the maximum effect was commonly attained within
one or two days at the most. The fact . . . that practice or
habituation to these changes greatly reduces the time required was
clearly shown. . . . Certain specimens, after several changes of
background, were found to adapt themselves, in almost full meas-
ure, to one of these within a fraction of a minute.

Some of the adaptations were so perfect that the flatfish was
all but indistinguishable from the background ; and the spots on
the skin became smaller when the diameter of spots on the back-
ground was reduced by less than a centimeter. Naturally, the
imitation was by no means equally successful against all back-
grounds. "Fixed morphological conditions" prevented the re-
production of perfect squares, triangles, circles, etc. It was, how-
ever, so exceedingly plastic and sensitive that "the notion that
the fish is limited to a few stereotyped responses, representing
the most familiar types of habitat, must be rejected at once."

The discoveries of greatest interest are two : first, the imitation
is accomplished through the flatfish's eyes; and, secondly, only a
part of the fish's visual field is involved in the process — the creature
does not imitate everything it sees. The first discovery might have
been made by pure analysis. For, were the eye not the instrument
of imitation, then we should have to assume that the light reflected
up from the background acted directly on the skin. Against this,
however, there are two objections : first, the flatfish's belly, which
directly receives the reflected light while the fish is swimming above
the sea-bottom, is not sensitive — only the back, turned away from the
reflected light, takes on the patterns; and, secondly, as Summer
points out, "it is impossible to see how responses to a pattern
could be brought about through any organs except the eyes, for
these alone are provided with the lenses necessary for the produc-
tion of images." The reader who suspects the analytical method,
however, may ignore these proofs.1 For Sumner has demon-

1 The possibility that tactile stimuli may produce the effect cannot be absolutely
eliminated by pure analysis. Sumner has put it out of court, though, by the
simple experiment of putting fish on glass bottoms, the under side of which waa
painted with patterns. The imitative reaction was exactly as quick and as sure as
ever.

400 IMPLICATIONS OF BIOLOGY

strated the matter by cauterizing the flatfish's eyes with silver
nitrate, blindfolding them, or blinding them completely. So
treated, the animal ceased altogether to imitate the background
pattern, and its hue reverted to an even, dark shade "representing
more nearly the resting state of the chromatophores."

The second discovery grew out of a series of experiments in
which the walls of the tank were variously colored and patterned.

In the case of the Rhomboidichthys, . . . that part of the bottom
immediately surrounding the fish appeared to be the one chiefly
effective. . . . The influence of the vertical walls of the vessel
commonly seemed to be a subordinate one, even in cases where the
fish was so large that it covered a considerable fraction of the bottom
and was obliged to lie constantly with its eyes close to one side or
another of the jar. . . . What the fish saw directly overhead . . .
seemed to exert a negligible influence upon the color pattern.

So much for the chief facts. Be it noted, first of all, that the
imitative reaction is not always seen by the flatfish, and that
Sumner has shown that the animal can adapt while its entire body
(except for the eyes) is buried hi sand or completely masked with
a cloth or deeply stained. This renders it "highly improbable
that any direct visual comparison on the part of the fish between
its own body surface and the surrounding background is an es-
sential factor hi the production of these changes." In other
words, 'consciousness' (in whatever sense the term be used) is not
an instrument in making one space pattern match another.

If it is not, how can one continue to hold the old psychological
doctrine that the arrangement of space-elements (or non-spatial
elements) into forms, patterns, or perspective orders is brought
about in any degree by the cognitive process, or by the 'associa-
tion of ideas' ? As with the imitative reflex among human beings,
so here. The process is set up by a physical stimulus, and its
result either cannot or need not be perceived by the organism.
The correspondence, therefore, is not between the flatfish's percept
A and the same flatfish's percept B; it is between a stimulus (which
may or may not be perceived, for aught we yet know) and a chemi-
cal pattern which is the cause of a perception in an external observer.

TRANSLATION OF PERSPECTIVE 401

In other words, the flatfish is not imitating merely its own percept
A, but is doing so in such a manner that some other creature will
perceive the flatfish's skin as having, not merely the characteristics
of the fish's percept A, but as having the characteristics of the external
cause of perceiving A. To make this last point clear, consider the
second phase of the total process; namely, the spatial transformation.

The flatfish's eyes are very close to the sea-bottom, sometimes
only a centimeter or so above it — as when it buries its body. It
is while in this position that the eyes sense the shape, size, color,
and arrangement of the sand and pebbles. These objects are
therefore cast upon the retina in an extremely oblique perspective.
You may get the general effect by holding your eye close to your
desk and glancing across the latter. The foreshortening will be
at a maximum ; a few items in the foreground will loom up, while
the converging of the rest of the field will be rapid. But this is
not the scene which the flatfish reproduces on its skin. It depicts,
with its chromatophores, the color, shape, size, and pattern of the
material on the sea-bottom as this material would appear to an eye
whose line of direct vision was perpendicular to the plane of the sea-
bottom and a great enough distance from the sea-bottom so that the
units of the pattern could be seen without any appreciable perspective
distortion. Roughly speaking, the skin pattern closely resembles
that of the sea-bottom as the latter would appear to you if you
were looking directly down at it, and six inches or more away from it.

However this queer deed is accomplished, it certainly results in
a translation of one perspective into another perspective; and
this translation is precisely that which is deducible from Euclidean
space. The rate of reduction of relative sizes in the first perspec-
tive is a function (mathematical) of the distance between eye and
plane ; and this very same relation governs the chemisms in the
chromatophores and also the rearranging of the latter.1 Inasmuch

1 Sumner has found that the plane in which a given surface lies with relation

to the flatfish sometimes determines whether or not it shall be effective in calling

forth a given change. It is not certain, he adds, that this influence is decided by

" purely quantitative relations within the visual field." Should later experiments

2D

402 IMPLICATIONS OF BIOLOGY

as this translation is accomplished without the fish's seeing what
it is doing to its own skin, we must at least conclude that somehow
the perspective relations are so thoroughly 'in' the space which
the fish sees around it that they can cause other perspective rela-
tions just as truly as one chemical relation causes another. And
just as one chemical relation causes another, without the assistance
of any psychical act, so too with perspective relations : they are
not set up by the 'association of ideas/ nor by an 'a priori syn-
thesis.' They are physical, no less than weight is, and absolutely
non-mental in the sense that they are not constituted by any
psychical process. The full interpretation of this particular mat-
ter lies beyond formal analysis. But it is not improper to suggest
that, if perspective is a mere physical relation in a lower animal's
adjustment, it is probably the same in human perceptions. At all
events, no mere introspective difficulties can now force us to as-
sume that some mysterious psychical 'association of ideas' or
'local signs' manufacture, out of non-physical elements, the space
relations which we see. Should this statement be taken in full
earnest, the entire British and Kantian psychology, together with
all modern disguised variations, would have to be discarded.

And now for the third phase of the process, namely selection.
The flatfish adapts its skin pattern only to the sea-bottom, in
normal life. The animal notices objects above the bottom and
even directly overhead; it follows such with its eyes and moves
toward or away from them. But the very same pattern which sets
up the pigment reactions when it is underneath the fish has absolutely
no effect when above it.

A plate of opaque white glass, of the same size as the bottom
of the tank, was covered with small, irregular blotches of black

paint The three specimens used in this experiment had

all been unmistakably influenced by this spotted plate when this
was placed beneath them, assuming a much blotched appearance.
Upon the removal of the plate from beneath them, they had re-

fully confirm the non-quantitative character of it, we should have to admit
what I have above advanced, namely, that directions are precisely as objective
and efficient as material things are.

SELECTIVE ATTENTION 403

turned to a nearly unspotted condition. The spotted plate . . .
was next inserted above the fishes (under the surface of the water,
of course). The plate . . . was brightly lighted by the mirror
below. That the fishes could see this spotted surface cannot be
doubted. Nevertheless, not one of the specimens showed any appre-
ciable influence, even after several days. Return of the spotted plate
to the bottom of the tank, beneath the fishes, resulted in each case in
a resumption of the blotched condition within a few hours at most.1

Descriptively, then, the perceiving function is not constructive
or transformative, but merely selective. Certain important ele-
ments in the environment are attended to, to the exclusion of
others, when and only when it is a question of adapting the skin
pattern. The other elements are seen, but they are ignored for
this particular reaction. Unquestionably, the flatfish has a gen-
uine 'field of attention' wholly distinct from the visual field.
But the items entering into the field of attention are not trans-
formed therewith ; for they all pass over unchanged into the skin
of the fish. I see no escape from the conclusion that at least one
of the primary functions of psychic reactions is to select, reject, and
direct certain environmental characters with reference to certain
other functions (such as nutrition, protection, locomotion, etc.).
It seems equally sure that selection and reaction to the selected
character does not necessarily alter the latter. In other words,
adaptation of agent to environment does not have to involve a
qualitative change in more than one of the two relata. And, in
the special case of selective attention, this operation does not mod-
ify the essential characters of the stimuli selected for response.2

1 Italics mine.

2 The remarkable facts Sumner has brought to light may not raise the difficulties
which their discoverer fears. I do not find it difficult, as he does, " to conceive of a
nervous mechanism competent to bring about such changes." Is it not quite
probable that we have to do here with an exceedingly simple, rather than a mysteri-
ously complex structure and function ? Instead of being an elaborate photochemi-
cal process which begins in the retina and undergoes a large number of incompre-
hensible transformations on the tract to the pigment cells, may the patterning
not be effected by the direct conduction of untransformed ether waves to the
chromatophores ? To-day this hypothesis is not so extravagant as it was before
Sherrington and others demonstrated experimentally that the periodicity of nervous

404 IMPLICATIONS OF BIOLOGY

We have still to consider a case of conduction which is even
more illuminating than the untransformed conduction of stimuli
through the flatfish. In the latter we found peripheral conditions
flowing on through the organism ; and we were justified in infer-
ring that, whatever the sensory nervous system of the animal ac-
complished, it did not alter all of the stimulus characters. We
were not assured, however, that the peripheral sense organs did
not add some psychic quality to the received stimulus; and it
remained at least conceivable that this additional tone qualified
whatever may have appeared in the animal's consciousness and
perhaps have been the necessary ingredient in the dynamic complex
which later regulated the imitative adjustments. In other words,
while the sensory nerves conduct some qualities, they may, so far as
the case in question testifies, add to the latter some unique speci-
ficity ; and this might be of the sort which anti-realistic thinkers
would describe as 'physical' or 'mental' or 'non-physical.'

But there is another well-established fact which demonstrates
that, even though sensory nerves (or structures) do add some-
thing to stimuli, that which is added is a character which can be
propagated through inorganic spaces, precisely as light or elec-
tricity can be ; and hence, that it is, in some sense, spatial or physical.
I allude now to the experiments of Nemec, proving that, in plants,
stimuli that have been taken up by leaf and stem tips can pass
through dead tissue and induce in the organism the same reaction
as though the stimuli had been transmitted through living tissue.1
The following summary of Nemec's research is abridged from
Macdougall's account of it.2

impulses corresponds to the periodicity of serially repeated stimuli, at least up to
rates of 500 per second. If we suppose that this correspondence holds for all cases,
even those of ether wave shocks, we then have a nerve impulse at the pigment
cells which is, in pattern and in photo-chemical efficiency, identical with the stim-
ulus pattern.

1 Nemec, B., Reizleitung und die Reizleitenden Strukturen bei den Pflanzen,
Jena, 1901.

» Macdougall, D. T., Practical Text-Book of Plant Physiology, New York, 1901,
16, 17.

TRANSFORMATION 405

A specimen of Mimosa was employed. A section of its stem
was killed by pouring boiling water on it for three periods of five
minutes each. Harsh stimuli were then administered to leaves
above the injured section. Some leaves were cut, and others were
burned. The reactions were of the usual sort, in spite of the fact
that the shock had to be transmitted through dead cells before
the reactions could occur.

Now, one might say that the stimuli were conveyed through the
dead section by the water in the tube-like cells which are supposed
to conduct impulses consisting of hydrostatic disturbances of the
fluid they contain. But this does not seem to be the case, inas-
much as the reaction of the injured Mimosa takes place also when
the stem is dessicated. But suppose that, even in this latter in-
stance, there is a microscopic amount of fluid present in the in-
jured region, and that it is by means of this fluid that the reaction
is regulated. Would this not still force us to infer that the speci-
ficity of the stimulus and of the reaction is not intracellular, but
is somehow a property of the fluid; and a property, moreover,
which can flow along through dead matter, just as an electric cur-
rent can ?

To be sure, the experiment does not prove that the specificity
of the stimulus is not conveyed through dead tissue, say, by some
unknown microorganisms in the fluid, and that these micro-
organisms infuse into the received stimulus some mysterious
psychical, non-physical nature. I leave the reader free, however,
to attach what importance he will to this conceivable state of
affairs. For myself, it does not seem worth a moment's consid-
eration, in the light of the other facts we know about afferent
processes.

d. Transformation. — Type Al. This has just been described
in the case of the flatfish.

Type A 2. — This reaction is the most difficult to describe, because
it is either going on simultaneously with a selective process or else
operates with a stimulus of which we know so little that we cannot

406 IMPLICATIONS OF BIOLOGY

be sure that there is a genuine transformation. The former case
we find in the digestive series ; the latter case appears in the so-
called 'lower senses,' touch, taste, and smell. Especially in the
tactile sensation do we seem to have a qualitative change of the
profoundest character. I mean, the stimulus is simple pressure,
in the thoroughly physical sense of this term; but the specific
quale resulting from that pressure cannot readily be thought of or
perceptually analyzed into precisely what the physicistmeans by pres-
sure. There would appear, therefore, to be here some inner change
of the stimulus effected by the reaction to it. Pure description can
carry me not an inch beyond this view. I know of no instances
which give us evidence that pressure is homogeneously propa-
gated through the nervous system, as light and air waves appear
to be.1 Indeed, I do not understand what that performance would
be. So I shall say that, so far as description can now carry us, the
stimuli of odor, taste, smell, and other 'lower' senses are transformed
by the reactions to them. This does not mean, of course, that it is
the sensing, as such, that changes them ; it only means that, in con-

1 Difficult as a matter of pure description, I mean. I have a hypothesis about
this very difficulty, and the hypothesis involves an explanation of the striking
differences between the tactile, olfactory, and gustatory characters and those of the
'higher senses.' Briefly, it assumes that the 'lower' sense operations are ' internal '
to each stimulus unit, whereas those operations of the ' higher ' senses are 'external '
to their stimulus units. By ' internal,' I here mean that the ' raw ' stimulus actually
invades the sensing organ ; by ' external,' I mean that it does not. When it does in-
vade it, the total organ is taken up with its reacting to the single stimulus and so,
sensing nothing else, does not sense the difference of the momentary stimulus from
others. Lacking every such differentia, the stimulus is specific but not explicitly
qualified. To speak the language of logic, the quale is a term stripped of all its re-
lations. Hence our inability to classify and order the ' lower ' qualia; hence
their inability to compose themselves into forms analogous to a musical chord
or a space-color pattern.

All this and its extensive presuppositions, however, are not formal analysis. I
hint at them only to prevent a critic from supposing that, when I admit the failure of
pure description to state the relation between stimulus and reaction in the ' lower '
senses, I am confessing that a realistic view breaks down. This latter does not
depend upon the universal possibility of formal analysis. To say that it did would
be to presuppose that no genuine simples are given in experience, and also that
every given complex must be equally clear as to its terms as it ia in its relations.

TRANSFORMATION 407

nection with the sensory reaction, some process does. The process
that does may bear the same general relation to the sensory process
that the masticating of food bears to the digestive selection of
food elements. Indeed, to lapse for an instant from pure descrip-
tion, there are many good reasons for supposing that it is just
this relation which does obtain; and there are some grounds for
conjecturing that most, if not all, cases of transformation precede
selective processes and serve only to facilitate selection. Surely,
all the clear instances of transformation, to which we shall now
turn, are so to be described.

The two great, unambiguous types of transformation are res-
piration and digestion. Here the stimuli are actually seized
upon and made over in toto. In toto, be it repeated, but not in
partibus. That is, the received complex is, as a complex, increased
by the addition of new elements (as saliva, bile, etc.) which alter
the structure of the complex. The new complex is so formed that
certain of its elements can now be withdrawn from it more read-
ily; and they are so withdrawn. (Once more let me warn the
reader that this language does not imply or presuppose finalistic
processes.) After the elements have been withdrawn and appro-
priated by the organism, the remainder of the complex (including
some original and some acquired constituents, in mixed measure)
is cast out of the organism. The rejected part is cut off from stimu-
lating and regulating the organism, in its further reactions to the
selected part of the complex.

This general description is enough to make clear the broader
character of the digestive and respiratory transformations. Were
we conducting a biological canvass, we should now have to ask in
which other operations the same type of transformation occurs.
That is not our present duty ; but I should like to point out one
peculiar instance. In so far as pure description guides me, I find
pleasure and pain to be precisely such additive transformations,

Formal analysis is only one of many methods of discerning reals. Formal analysis
may be just as incompetent to deal with some problems as deduction is useless in
one's endeavor to decide which of two paintings is the more beautiful.

408 IMPLICATIONS OF BIOLOGY

serving to facilitate a later selection of some element. This de-
scription does not tell the whole truth about the situation, but, so
far as it goes, it seems to fit.

I am directly aware of two circumstances in many algedonic
reactions : first, that the ' external ' stimulus is colored or suffused
with the feeling tone, which is felt as not belonging to it in the
same thorough manner in which its specific quality does; and,
secondly, that, in the succeeding reactions to this complex, the feeling
tone is sloughed off and the organism operates solely with the ' external'
character. For instance, I hear a sound which is very pleasant;
and the pleasantness of it 'makes' me attend to it exclusively. I
select the sound from a large complex of auditory stimuli; but
after I have selected it, I react to the objective character of it
altogether (or, at least, predominantly). My motor adjustments
are to its rhythms, not to the pleasure ; my 'associations ' (reveries)
are about its specific character, its relation to other melodies, other
places where I have heard melodies, its meaning, etc. In other
words, the pleasure initially transforms the music ; but the conse-
quence of its doing so is that it facilitates the selection of the music, and
after the selection the pleasure tone is not reacted to, but at most
only accompanies or coexists with the selected music. So far as
pure description can speak, then, the pleasure tone is, to the mental
reactions, exactly what saliva is to the digestive reactions; a
transformer by addition, serving to facilitate a later selection of
something else.

Pleasure and pain, then, are intraorganic additives. Like saliva,
they are not an original part of the incoming stimulus nor yet
mere products of the assimilating process, but rather a prelimi-
nary aid to a selection and a later adjustment. They mingle with
the received matter, and genuinely qualify it in such a manner that
the organic reaction to the matter is heightened. Thus, they most
remarkably resemble the opsonins of the blood, whose specific
function it is to 'flavor' bacteria so that the leucocytes will ab-
sorb the latter. This resemblance, I admit, may be quite acci-
dental and without much significance; but it is worth some in-

RESISTANCE 409

quiry, for it would be a discovery of no small moment, were we to
find that blood reactions and algedonic reactions are generically
related.

Type B. — I have not been able to put my finger upon a case
of such transformation. Always the instances which at first seem
to fulfill the conditions of this type prove, upon inspection, to be
only complex and obscure instances of either Type Al or A2.

e. Resistances. — These are structural or functional or both. For
our purposes, they need not be so distinguished, though. It is
here important only to indicate the type of relation between the
stimulus and the resistance reaction. Now, there are at least two
and almost certainly three or more radically different sub-types
of this relation : (i) resistance by destruction of the stimulus (or
an element of it) ; (M) resistance by interpolation of insensitive
material between reagent and stimulus; and (Hi) resistance by
suspense of collateral functions. Strict accuracy might lead one
to add a fourth sub-type, namely resistance by simple rejection
of the stimulus ; as in the case of resisting attack by a wild beast.
But, for obvious reasons, we need not concern ourselves here with
this instance.

(i) Resistance by Destruction of Stimulus. — Specimens of this
are very numerous. I mention only the blood resistances, be-
cause they bring out the important features of the type.

a. Antitoxins of the Blood. — The specific antitoxic reaction of
the blood does not act upon the bacteria which have produced the
toxin. It is simply a process of neutralizing the toxin. The
diphtheria bacillus, the tetanus bacillus, and other varieties gene-
rate poisons injurious to the organism they inhabit. These poisons
set up in the organism a counter-process whereby a chemical is
generated which combines with the toxin so as to neutralize it.
This antitoxin is absolutely harmless to the bacteria themselves,
which thrive in it jauntily. It is evident, therefore, that the bac-
teria are not the stimuli of this antitoxic reaction. The latter is
set up and directed toward one particular effect of the bacteria. In

410 IMPLICATIONS OF BIOLOGY

short, the organism here reacts to only one feature of the 'total
situation.'

ft. Bacteriolysins and Bactericides. — These are two other varie-
ties of organic chemicals generated by the organism infected. They
possess the contrary power of the antitoxins. The latter neutral-
ize the toxins produced by the bacteria, but do not injure the
bacteria. The bacteriolysins and bactericides, on the other hand,
do not neutralize the toxins at all, but they do kill and consume
the bacteria. Here again, then, we have a reaction directed to-
ward a particular feature of a 'total situation' and affecting only
that feature.

Of such chemical reactions the number is legion ; and not only
is each one directed toward some element in a stimulating complex,
but also it is efficient only with respect to that very element. The
tetanus antitoxin neutralizes only the toxin of tetanus bacteria,
not that of diphtheria or any other infection. The typhoid bac-
teriolysin avails naught against the malaria bacteria. And so on.

Another observation which will bear upon the philosophical
issue we shall raise. At least five types of blood reactions are pro-
duced simultaneously in case of infection : the three above named
and also agglutinators and opsonins. But, though produced as
if by one reaction, they do not form an 'organic whole' in the
Hegelian sense of this misleading phrase. That is, they are not of
such a character that their nature is constituted by their inter-
relations. We have of this a perfect empirical proof, in the fact
that they can be separated by the physiological chemist without
losing their specific efficiencies. The antitoxic serum can be with-
drawn from the blood of a horse, and, apart from the other chemi-
cals of the blood, still neutralize its own proper toxin.

The question whether such resistance chemicals depend upon
a natural reagent at all — that is, whether they are caused exclu-
sively by an organic process — has not yet been conclusively an-
swered. The trend of evidence, however, is strongly toward a
negative answer. Organic secretions have nothing unique or mys-
terious in their make-up. They can doubtless be manufactured

RESISTANCE BY INTERPOLATION 411

in any laboratory, as soon as their chemistry is better understood.
When the first one is so produced, biologists will see more clearly
that the peculiar 'unity' of life is not to be looked for in organic
chemistry, but rather in the dynamic relation between stimulus
and the response to its specificity.

(ii) Resistance by Interpolation. — Here are to be included all
operations like callousing, thickening of fur, pigmentation under
sunlight, etc. Suppose my hands, now soft and thin-skinned, are
put to handling rough boards. At first, the skin is scratched and
worn through and blistered. But soon it thickens so much that
repetitions of the previous stimulation no longer reach the sensory
nerves and the subcutaneous tissues with the same violence. Now
the immediate stimulus is a series of impacts, and these are halted.
But this halting does not constitute the reaction of which we are
speaking. The callousing is a wholly different operation. It
succeeds the absorption of the immediate impact by a consider-
able period ; and — what is infinitely more significant — it
is not directed to the stimulus which induced it (for that stimulus
has long since passed), but is directed to the path of the stimulus,
and so indirectly to future stimuli. And the effect of the reaction
is not to prevent its own initiating stimulus from doing damage to
the organism, but to prevent later stimuli from so doing. I claim,
therefore, that, as a matter of pure description, we ought not call
the callousing a reaction to the stimulus initiating it, unless we
peculiarly limit the meaning of the preposition 'to.' If Ho'
indicates the direction of the reaction, the preposition is certainly
out of place in the above context.

We may now state more precisely the manner in which the relata
are and are not reciprocally altered in this type of reaction. The
immediate stimulus is absorbed, not by the reaction, but by the
organic stuff that intercepts it. The reaction grows out of this
stuff, somehow, and operates, not so as to change the environ-
ment (e.g. the boards that are being handled, or the handling of
them), but only so as to cut off the impacts from the sensitive or-
ganism. This does not change the impacts, in the sense that they

412 IMPLICATIONS OF BIOLOGY

are any different, as objective, external events. A board, for in-
stance, is equally rough before and after the callousing reaction
has set up. It weighs just as much, whether the lumberman's
palms are soft or hard. The inertia and momentum, as it slides
off the lumber wagon, are not altered by the reaction.

How, then, has anything been accomplished by the reaction?
The answer carries us back to the difficulty we alluded to at the
beginning of this essay; namely, the demarcation of organism from
environment. If we define the organism as that which initiates
reactions to stimuli, then large tracts of the body are not organic,
in this narrower sense. The callouses are not, for instance. They
are products of a reaction, not the initiators. Should we employ
that definition, we ought to describe them as elements inserted
into the environment by the organism, as insulators. As objects,
they bear the same relation to the biological 'situation' that the
glass insulators of a telephone line bear to the telephoning.
With respect to the latter process, the insulators are a true part
of the environment; they are not a phase of the telephoning -pro-
cess, but only a means of preventing the earth from disturbing
that process. They neither modify the elements in the telephonic
series nor the earth, but rather forestall such modifications.

(in) Resistance by Suspense of Collateral Functions. — This process
appears in that all-important but little understood operation of
mental attention, which Ebbinghaus aptly describes as 'eine
rechte Verlegenheit der Psychologic.' If the reader will bear in
mind that I am not here engaged in presenting a theory about
attention, but am only describing what is open to formal analysis,
he need not be offended by the remark that attention not only in-
volves a certain facilitation whereby the observer smooths the way
for stimuli of the chosen type, but it also involves cutting out all
other stimuli — or at least offering them a heightened resistance.

As to the precise nature of this double performance, psycholo-
gists know little or nothing. They conjecture that the facilitation
may be accomplished by a drainage of nervous energy into the
appropriate sensory centers from the muscles which are accommo-

SUSPENSE OF COLLATERAL FUNCTIONS 413

dated specifically to the stimuli that are to be attended to. But
this strikes me as unillum mating ; for the accommodation itself is
certainly a part of the attending act, indeed may be the facilita-
tion and not the cause of it. And, in any case, there is little or no
physiological evidence of such a drainage, and there is even con-
siderable doubt as to the existence of sensory centers, especially
of the sort which are so highly differentiated that one receives only
the pitch of tone and another only the timbre, one the form of a
visible object and another the color. Because of such uncertain-
ties, we must shun every exact theory about attention and con-
tent ourselves with pointing out certain respects in which its
specific resistances — whatever they may be — contrast with the
types above described.

Now, it is safe to say that the inhibited stimuli are not destroyed,
as bacteria are by the bactericidal reaction. For we commonly
become aware of events which we did not cognize while they were
happening because we were engrossed with some other affairs.
At the close of an exciting chapter in a novel, we suddenly sense
for the first time that the clock struck midnight some minutes ago.
Of course, this could not occur if the original stimulus had been
annihilated somewhere along the afferent tract.

Again, there cannot be any interpolation of insensitive struc-
ture, as in callousing or protective pigmentation. Not only is all
evidence of such a process lacking ; the latter is inconceivable. No
systems of screens can be imagined which would shut out or inter-
fere strongly with all but any one kind of that incomputable multi-
tude of stimuli which assail the sense organs continually. How,
for example, could there be devised a material filter that would ad-
mit to the cortex only the meaning of words and not their printed
form (a separation which is readily made, in attention) or conversely
only the printed form and not their contextual significance ? l To

1 This phenomenon is less familiar and, so far as I know, has not been studied.
It sometimes appears in highly expert proofreaders, who have trained themselves
so well to note only typography that often they cannot state even vaguely the con-
tent of what they have just read.

414 IMPLICATIONS OF BIOLOGY

put the question is to answer it. The filter would have to be of
infinite complexity and instantaneous plasticity, to manage the
swift shifts of attention in everyday life.

Now, if the stimulus is not destroyed, and if it is not excluded
from the attention field by an interpolated structure, then there
remains only one member of the biological triad to perform the
trick; and that is the reagent itself. It must suspend certain
operations. It must simply stop sensing certain data in certain
contexts. How this is done, we do not know; nor need we, so
long as we are merely analyzing observable events in terms of
their observed differences. This much may be anticipated, how-
ever; the functions which are suspended are not mere spatio-
temporal movements, for they do not effect the displacing of the
inhibited stimuli in time or in space, but only their displacing
from the cognitive field. Doubtless they involve spatio-tem-
poral operations; but these do not constitute, of themselves, the
specific resistance. If this statement is obscure, the reader may
ignore it for the present.

Ill

RESULTS OF THIS FORMAL ANALYSIS

At least four highly significant facts have emerged from the
above descriptions.

1. Reaction ('response') is not a single type, except in so
far as every case of reaction involves the same types of complexes
as its terms, namely, 'agent' and environment. With respect to
the relations between these terms, there are many heterogeneous
reactions which cannot be reduced to any one pattern.

2. The relations between the terms of the biological 'situa-
tion' are not 'internal,' in any observed case. Transformations
which seem to produce new, unique qualities do so only in the
sense that they establish new complexes which, with respect to
their complexity alone, are novel. Primordial characters do not
seem to be created by any organic change which we can observe.

RESULTS OF THE ANALYSIS 415

3. Reactions are not to 'total situations,' but to phases or parts
or elements of such. This narrowing of activity does not always
entail a change in the environment nor the manufacture of synthetic
products in the organism.

4. Geometrical, mathematical, and other relations are gen-
uine stimuli, in the very same sense that material complexes are.
Hence, for formal analysis, they are not products of the cognitive
reaction, but the producers of it and many other kinds.

A few comments on each of these points.

1. This must be urged upon idealistic biologists and also upon
some biological pragmatists. These investigators have tended
to construe all reactions as fundamentally alike, not only in re-
spect to the terms involved, but also in respect to the type of re-
lation between these. This relation has been called a transform-
ative one, and the adjective has not been severely analyzed.

The biological pragmatist, for instance, opposes radical
realism on the ground that adaptation involves some trans-
formation of the environmental character which stimulates the or-
ganism and to which the organism responds. He admits that
there is an external and independent environment, and just so
far he is one with the realist. But he denies that there is anything
like correspondence between the environment as known and the
environment as unknown. For, says he, knowing is a response to
a situation; and wherever such processes occur, there is a
modification of both the stimulus and the recipient organism.
Therefore, the world, as known, is a peculiar system of 'contents,'
which do not exist as qualities of the external order, but are never-
theless genuine functions of the two variables, stimulus and re-
action processes. In this manner the biological pragmatist es-
capes epiphenomenalism and subjectivism, and he seems to have
struck a happy compromise between the radical idealist and the
radical realist. With the idealist, he maintains that there is
some kind of synthesis involved in experience, and that this syn-
thesis eventuates in a 'content' which is neither pure object nor

416 IMPLICATIONS OF BIOLOGY

pure subject. With the realist, on the other hand, he believes
that the object system (that is, what the biologist calls the environ-
ment) is in no sense constituted by its being known. So it comes
that he finds himself unscathed by the realistic attacks upon the
idealistic doctrine that esse = percipi. Also he is not put to it
to clear up the difficulties in the theory of consciousness and the
postulate of 'external relations/ as the realist is. All in all, then,
his lot to-day seems a most happy one.

If, however, we inspect his hypothesis of stimulus and reaction,
we find that he has placed several restrictions upon the facts which
the descriptions we have been giving will not allow. This appears
most clearly in Dewey's account of the matter. Dewey believes
himself to be a realist with regard to percepts.1 But he finds reason
to maintain that the perceptive reaction is transformative and
constitutive.

The seen light is not in relation to an organism. We may speak,
if we will or if we must, of the relation of vibrations of the ether
to the eye-function ; but we 'cannot speak, without making non-
sense of the relation of the perceptual light to an eye, or an eye-
activity. For the joint efficiencies of the eye-activity and of the vi-
brations condition the seen light.2

As I understand this passage, it means that the 'joint efficien-
cies' bring about either an A2 or a B type of transformation of
the stimulus.3 If the eye-activity simply picked out and con-
ducted certain kinds of light to a center, and if also the seeing of
the light consisted merely in the latter's being related to certain
other physical entities (perhaps as elements are, in the cross
section of some process), then there would be no nonsense in
speaking of the relation of the perceived light to the eye, or to the
eye-activity. The eye-activity would bear somewhat the same
relation to the perceptual light which a camera bears to the light it
has rearranged. This relation might be statable very precisely ;

1 Cf. especially his Brief Studies in Realism. /. of Phil, Psychol., etc., 8, 393 ff.
* Ibid., 396, note. Last italics mine.
» Cf. 386 above.

THE PERCEPTIVE REACTION 417

and, I shall venture to suggest later, that the general type of it is
strikingly like a familiar geometrical relation.

To deny that such a relation obtains is justifiable only on the
assumption that the eye and the ether in combination somehow
constitute the seen light, after the manner of a Type B transforma-
tion. Were this the case, it would be futile to search for a rela-
tion between the relation of eye to ether and the eye. Such would
be a Bradleyish procedure, precisely like that of hunting for the
hidden relation between A and r, in the complex, ArB. Now, to
condemn our describing the seen light as related to the organism
is to presuppose that the seen light is not an entity in relation, but
the relating of two entities. It is this view which Dewey ex-
plicitly holds, not on general metaphysical grounds, but for purely
empirical reasons.

This, I believe, cannot be maintained against the evidence of
biology. If there were no better facts to array, the imitative
adaptation of the flatfish would prove conclusively that, at least
in some cases, the eye activity does not condition the specific
light-character of ether vibrations, but only the distribution and
employment of these. To be sure, one might still say that the
organism conditions the seen light ; but exact description quickly
reveals the danger in such a statement, for it is not the lumin-
osity, as such, but some other peculiarities (especially the geomet-
rical) which are conditioned.

The most effective facts to cite here, though, are those which we
know directly and which have been long studied ; namely, the facts
of least perceptible differences. In these we have proof that the
joint efficiencies of eye activity and ether vibrations do not condi-
tion the seen light. Unlike instances drawn from the behavior of
lower animals, the psychological data to be considered forestalls
one common and difficult retort of the internalist. "If," he often
says, "you but knew A in all its inwardness, you would discover
that every change in r involves a change in A (and B). A tree
may seem to be unmodified by being perceived ; but that is only
because we know not all that is happening to it." However

418 IMPLICATIONS OF BIOLOGY

forcible this may be when we are considering physical objects at
large, it fails completely when we are debating about the immediate
nature of the percept alleged to be conditioned by the organic and
the physical terms of which it is a relation. For now we know in
its entirety the subject of discussion; we are talking about the
percept, qua percept.

I gaze at a pane of ground glass which is illuminated from be-
hind. An experimenter slowly increases the number of standard
candles which he allows to shine upon the glass from a certain dis-
tance. If the distance is great enough, he may light up twenty
candles before I perceive an increased brightness upon the ground
glass. In terms of relations and relata, what has happened? A
stimulus has been modified twenty times, and each time a new
manner (by another increment), but the percept has not changed
at all. If, then, it be admitted that there is any relation whatso-
ever between the stimulus and the percept, it must be granted also
that the latter at least is not constituted by that relation.

At first thought, one might be tempted to escape this conclusion
by arguing that the changes in the stimulus have been 'absorbed'
or 'overcome' somewhere along the physiological tract that is
involved in the perceiving process. But this avails not at all, so
long as one adheres consistently to the theory of internal relations.
For let these physiological factors be m, n, o, p, q, r, . . . in the
first instance, before the stimulus has been altered. Then, the com-
plex S-m-n-o-p-q-r- . . . C constitutes C (the perceived brightness).
That is, the relation of C to r, of r to q, of q to p, of p to o, etc. back
to S makes C what C is. But if S changes its relation to m (as it
must if S is increased by twenty candle power), then m is altered ;
for m, like all other entities, is modified intrinsically by its relations
to other entities. And, if m is altered, then its relation to n is,
and consequently n too is inwardly changed. And this alteration
continues along the entire chain ; hence, whatever the last member
of the complex is prior to C, it must be different, hence its relation
to C is; and hence C itself is. Resistance, absorption, adapta-
tion cannot occur, inasmuch as they involve the overcoming of

PERCEPTION IS NON-CONSTITUTIVE 419

changes and, at the same time, the keeping of some entity (char-
acter, function, etc.) unchanged, in the affected system.

It is not necessary to point out the series of transformations in
the above-named series of relations, in order to prove that the
internal theory of relations is false here. We shall score no less
surely, by indicating what happens when the twenty added candles
are extinguished one by one. After the nineteenth has been put
out, the original stimulus no longer exists in any sense compatible
with the hypothesis of internal relations. There was a light of n
units power; there is one of n— 19 units. Nineteen relata have
dropped out of existence ; each was related to the percept ; hence
the latter has lost nineteen relations. But not even nineteen
losses alter it inwardly.1

From this there is no escape. It will do no good to suggest that
maybe there are minute, unnoticed changes in the percept during
the changing of the stimulus. Grant that there are; the diffi-
culty for the internalist is not a shade less formidable. For a
minute change in a phase or element of a complex inevitably modi-
fies the complex, as a whole. But what is the percept as a whole,
if not just that which we perceive, namely, a certain brightness?
This whole, however, is not changed at all. And if we do not
regard the percept, qua percept, as the whole, what in heaven's
name do we mean by the seen light, the heard sound, the felt smooth-
ness, and so on? In this particular debate, an appeal to 'subcon-
scious ' qualia would simply evade the issue.

In view of all this, I conclude that the discontinuous variation
of percept values during continuous (or more nearly continuous)
variations of stimulus proves empirically that the joint efficiencies
of the eye activity and the physical matter do not condition the
percept constitutively ; that is, the seen light or the heard sound
is not a new total complex of which stimulus and reaction are ele-
ments (Type A2 transformation), nor is it the dynamic relation
between stimulus and reagant and so constituted by the terms in

1 1 treat each candle as a unit relatum, simply for convenience. It would be no
less proper to speak of each ether wave as one.

420 IMPLICATIONS OF BIOLOGY

the manner that the relation between positive and negative electric
charges is constituted by these. (Alleged type B transforma-
tion.)

The difficulties which American biological pragmatists have
found with the concept of transformation and constitutive pro-
cesses can, I believe, be traced directly to the fact that, following
Dewey's lead, they have attempted to apply formal analysis, first
of all, to the ethical situation. Like Dewey, their first interest has
been in human conduct ; and, in describing its processes, they have
rendered a noteworthy service to philosophy. But the subject
matter offers peculiar obstacles. First of all, it is prodigiously
complex ; then, it is largely intangible ; and, finally, it is thickly
overlaid with traditions, prejudices, and suspected habits of speech
and action. The inevitable result is that for a long time the ana-
lyst must deal only with broad complexes. He must handle many
things as though they were one thing, and perhaps one thing as
though it were many. He is in the predicament opposite to that
of the mathematician, who quickly attains the simpler entities
relation of his subject matter, but approaches with ever decreas-
ing velocity the analysis of the highest real complexes.

In his early formal analysis of the moral situation,1 Dewey em-
ploys the concept of 'transformation' in a manner which, however
necessary at the time, was fraught with perils. It is worth criti-
cizing here because it has undoubtedly influenced many pragma-
tists' interpretations of biological facts, even though they may
have long since consciously modified it.

Take the passage in which Dewey analyzes the 'mediation of
impulses.' This process he describes as the "back-reference of an
experience to the impulse which induces it." And he maintains
that "the expression of every impulse stimulates other experiences,
and these react into the original impulse and modify it. This re-
action ... is the psychological basis of moral conduct." "Such
mediations constitute the meaning of the impulse — they are its
significance, its import." And, four pages further on: "We may

1 The Study of Ethics. Ann Arbor, 1897, 14, etc.

ARE IMPULSES TRANSFORMED? 421

recognize three degrees of completeness in this mediation. In the
most complete reaction, the original or natural impulse is com-
pletely transformed; it no longer exists in its first condition. Our
impulse to locomotion, for example, is entirely made over when
the reaction of other experiences into it is completed — when we
learn to walk. . . ."

Now, an impulse is an activity, and nothing else. Its character
is not at all derived from the particular elements which happen to
lie in the path of the activity, restraining or facilitating it. In
the 'complete transformation' of the locomotion impulse through
mediation, it appears to me that the impulse itself is not changed
much by those factors which come to regulate the leg muscles.
What is changed is only the path of discharge and the rhythm of
tonus. Now, so long as we do not clearly separate the activity
itself from the elements it immediately affects and those which
affect it, we shall call the total complex 'the impulse.' And,
calling it that, we shall have to say that it is changed in quality
whenever its rhythm or direction or rate of discharge is altered.
But this way of looking at it is demonstrably improper, inas-
much as many elements involved in the discharge-zone of the
impulse do not contribute to create the impulse. They are not
constituents of the activity, but only of the field of action. They
bear the same relation to the impulse that the iron filings in a mag-
netic field do to the magnetic force there. Their pattern is not
identical with the pattern of the pure force, and they may be with-
drawn without altering the latter in any respect. Had we time
now to pursue this analysis, we should find that 'mediation' al-
ters only a few factors; and that these are chiefly parts, not
the activity, but of the field. Enough has been suggested, though,
to show the dangers in the way of A2 hypotheses here.

If the danger is serious, when we are analyzing impulse, it be-
comes gigantic when we turn to perception. It is not very hard
to hold apart, in one's thinking, the impulse from the im-
pulse field, and to observe that the latter does not constitute
the former, nor vice versa. But how different the undertaking

422 IMPLICATIONS OF BIOLOGY

is with the perceptual reaction ! What here is the activity and
what the field of activity? And how are they related? What
changes the one, and what the other? These questions the bio-
logical pragmatist has not answered wrongly; he has not yet
faced them, so far as I know.

2. If former analysis is to be trusted, the doctrine of ' internal '
relations finds absolutely no confirmation hi biology. This is the
irony of fate; for those who accept the doctrine build it into a
metaphysic which they call ' organic. ' They name it after the
analogy of life because they are persuaded that life offers the most
convincing illustrations of entities being constituted by their rela-
tions, and of parts depending upon the wholes in which they stand.

In this belief, the internalists follow a vague popular opinion.
Men know that a human heart, torn from its natural lodging,
immediately ceases to beat; and that its former owner loses all
interest in daily affairs. Also, it is pretty well established that
you cannot spoil a man's stomach without spoiling his temper —
and so on. Like most easily formed opinions, however, this one
has truth enough for most practical purposes, but falls short of the
philosopher's full measure. For the evidence of modern inquiries
is all in the opposite direction. Indeed, experimental zoology and
surgery are performing feats which make one wonder whether
there is a single relation that is ' organic ' in the metaphysical
sense of this adjective. Note these two following discoveries:

A. Organic parts do not depend upon the whole in which they
naturally occur, except in an empty sense of the verb. — R. G. Harri-
son has grafted the anterior half of a frog of one species (Vires-
cens) on to the posterior half of a specimen of another species
(Palustris), and has successfully reared young frogs from the
combination. Each half preserved the peculiarities of its species,
and there was no trace of any mutual influence between the halves.1

A still more enlightening experiment was performed by E. Joest,
who grafted the halves of two different species of earthworms
together. There was no sign of reciprocal influence; so, to make

1 Reported by Morgan, T. H., Experimental Zoology, New York, 1907, 299 ff.

ORGANIC WHOLES AND PARTS 423

certain that this was not due to the fact that both specimens were
fully formed and thereby fixed in structure and function prior to
grafting, Joest later removed a part of one of the components.
Regeneration then set in. Now, the new part derived nourishment
from blood that came from both components; nevertheless it grew
true to its native part. The new material from the other part did
not alter it at all.

We need not multiply instances, of which there are many.
Analyze any one of them, and the metaphysician's dilemma
appears readily. To defend the internalistic theory, he must say
that each half of the grafted frog was changed in some unknown
manner — and then the heavy, heavy burden of proof rests upon
him — or else he must say that one half of a frog is just that half
of just that frog, only so long as it is a half of that frog; and the
very instant it is cut off, it is no longer a half of that frog. Should
he choose this alternative, he would be defining the part by its
participation hi the whole; and this, as Perry has shown above,1 is
either a tautology or a petitio principii.

B. Organic wholes do not depend upon their individual parts for
their total specific organic character. — As Perry has indicated, when a
given event can be caused or maintained by any one of several
other events or conditions, it is improper to speak of the former
as being dependent upon any one of the latter. Bearing this in
mind, we observe that an organism is, within wide limits, indepen-
dent of the particular parts composing it. Now, if you grant that
it is, within even a narrow range, you have ruled against the in-
ternalistic theory.

Many animals possess an astonishing regenerative power. Cut
off a fish's tail, and a new one grows ; and, oddly enough, the
larger the piece you remove, the more rapid the new growth —
within certain limits, of course. It need not concern us here how
this feat is performed. The significant feature of it is that those
very characteristics of the organism which, according to any internal
theory of relations, depend most completely upon the absolute integrity

1 Cf. 107 above.

424 IMPLICATIONS OF BIOLOGY

of the entire complex are the very characteristics which are least affected
by such subtraction of parts. That is, the uniqueness of a particu-
lar fish consists predominantly in that which we vaguely call its
' individual life. ' This embraces, among other unknown ele-
ments, those accumulations of behavior which we call either ' habit '
or ' experience.' Whatever else of the creature's nature may be
suspected of commonness, this much at least is unique and private,
in every legitimate sense of these adjectives. Now, so far is the
fish from losing its identity and its specific life through loss of its
taH, that the animal is able to construct a new organ which is, in
structure and function, identical with its predecessor. Certainly,
this seems to prove, in every manner except the empty one alluded
to, that whatever genuine 'unity' the organism possesses is not
dependent upon all its parts, and perhaps not upon any one, but
on a congeries. It is typically related to its physiological parts
as the ectoderm to the optic vesicle, in the frogs on which Lewis
has demonstrated the existence of a ' formative stimulus.' 1 This
investigator demonstrated that any part of the ectoderm in the
embryo will grow the lens of an eye, if a piece of the optic vesicle
is transplanted beneath it. Described in logical terminology, the
relation between vesicle and ectoderm is asymmetrical ; in terms
of physics, the process of formation is irreversible; and, meta-
physically, neither the active term nor the dynamic relation in-
volved depends properly upon the passive term of the complex.

3. It is not only the neo-Hegelians who insist that organic re-
actions involve the 'total situation' in which the organism finds
itself. This is held by not a few pragmatists, at least with respect
to psychic reactions, such as thinking and perceiving. Something
very much like it is also held by a few biologists, I believe, es-
pecially those who lay stress upon the organic process of 'learning
by experience.' Like many other inaccurate hypotheses, this one
contains a certain truth, insistence upon which is important, in an
age which is tempted to oversimplify organic activity. But, for

1 Lewis, H. W., Experimental Studies on the Development of the Eye in Am-
phibia, J. Exp. Zool., 1905, 2.

MATHEMATICAL STIMULI 425

all the merit of it, the hypothesis suffers from rough-and-readiness.
In a broad way, it covers a multitude of facts ; but, strictly con-
strued, it covers absolutely none.

In the more primitive reactions, only one feature of a single
stimulus is effective. Thus, in the light striking the cell wall of
Equisetum, in geotropisms, and in other familiar instances. Now,
we find this same behavior all the way up to the very highest
known reaction ; namely, that of mature selective attention and
directive action in man. To be sure, it is not equally prominent
in all types of reaction ; notably in normal perception, which is an
adjustment to a large number of simultaneous influences. But
this reaction is the least typical of all; and, whether typical or
not, it is not the only sort. For present purposes, this is all I
wish to show. One instance of reaction to only a part of a situa-
tion is sufficient to prove some kind of organic pluralism. The
problem of determining just what manner and degree of organic
pluralism obtains, is much too vast for these pages.

4. I am quite aware that, in asserting planes, angles, numbers,
ratios, and other such mathematical-geometrical characters to be
stimuli of the peripheral sensory organs, in precisely the same sense
that ether waves are, I am exposing myself to ridicule. For this
ignominy, though, I shall blame the facts themselves ; I see no other
way of describing them. My difficulty is not in the least miti-
gated or removed by the fact that nearly everybody except the
new realists finds it easy to describe them as 'intellectual ab-
stractions/ 'constructs,' 'shorthand expressions,' and the like.
In an earlier section of this essay, it was pointed out that, if we
have a right to call the length of an ether wave the cause of the
color of the sensed light, we must also declare the direction of the
wave the cause of such phenomena as heliotropisms and the spe-
cial modes of cell division these set up. With this description
we must make earnest. When we say that length is a cause of
color, we must mean by length just that which a geometer means
by it. Anything else is sheer quibbling.

Is it not clear, now, that the biologist who accepts the results

426 IMPLICATIONS OF BIOLOGY

of pure description is in a position to repudiate almost every mod-
ern philosopher's theory of knowledge? Unaided by esoteric
lore about the Ego, he can discard the doctrine of Pearson and
other positivists that geometrical entities are 'mere shorthand
signs ' for peculiarities of sensations ; and the doctrine of Driesch
that space is phenomenal (i.e. a form of experience, and not a
form of the physical world independent of experience) ; and the
doctrine of Bergson that mathematical-geometrical characters
are static artifacts created by the 'vital force.' In discarding all
such confused opinions, the biologist rips out the very founda-
tions of the stately theories which their champions have erected ;
for these have all built on the belief that space and time give rise
to paradoxes which can be solved only by pronouncing the whole
situation 'unreal' in some manner. This belief is incompatible
with the evidence upon which it is founded. If formal analysis is
employed in securing evidences of biological principles, then the
system within which these principles hold good cannot be other
than the one in which the evidences themselves occur. Those
thinkers who employ facts from one realm of existence, to prove
that there is another wholly different realm, inevitably commit
the fallacy of distinguishing indiscernibles.1 This fallacy I shall
now try to exhibit.

Pearson, Driesch, Bergson, and many others have all employed
formal analysis and have accepted as evidence facts discovered by
it, in spite of their broader conclusions against the validity of per-
ception and thinking. Especially Driesch and Bergson watch the
behavior of animals in space and time, and from their directions,
rates of motion, rhythms, and arrangement of space structures in
their own bodies or in external objects, they infer that there is a
'psychoid' or a 'vital force' at work, that it strives to accomplish
certain things, and is related in a certain manner to a transcendent
environment. Now, the contradiction in this procedure is palp-

1 The situation just described is one which gives rise to a self-refuting system.
Spaulding has amply dissected this latter; the ensuing pages will deal with a wholly
different phase of the matter.

THE FALLACY OF INDISCERNIBLES 427

able. But, though palpable, it ought to be analyzed minutely,
inasmuch as Driesch (and possibly Bergson) evidently believes
that its ill effects are forestalled by a metaphysical presupposition
viz. that, although all specific terms and relations in the phenom-
enal field are peculiar to this field, they are nevertheless typical
of terms and relations in the noumenal field.

For instance, in the latter, there is a dualism of some sort, one
member of which is that which we mean, in the phenomenal order,
by 'psychoid' or 'ego' or 'vital force/ and the other member of
which corresponds (mathematically) to the system which, in
phenomena, we call the 'environment.' Furthermore, the
'psychoid' or 'organism' of the noumenal order is confronted
with difficulties in its environment which it struggles to over-
come, and the manner in which it succeeds in so doing is typically
like that which the biologist observes in the phenomenal order.
Thus the flatfish has a noumenal existence, and in this trans-
cendental life it develops various activities of the same type as
those which we call 'appetite/ 'struggle for existence/ 'sex
impulse/ and so on. Some one of these activities involves the
process of bringing a part of its organism into the same type of
relation with a certain part of its environment as that which we
designate empirically as adaptation to the color and pattern of the
sea-bottom. And this relation obtains, not merely in general, but
in each moment and in each variety of adaptation. That is, when
the flatfish, having a certain blue-gray checker pattern on its back
as a consequence of resting upon a blue-gray checker sea-bottom,
shifts to a gray-brown sea-bottom of irregular design and there
soon develops on its back a gray-brown pattern of irregular de-
sign, the difference between the former and the latter noumenal
situation is of the same type as that between the blue-gray checker
and the irregular gray-brown pattern of the sea-bottoms. Never-
theless, the noumenal difference is not a difference between colors
and space forms, for these are only phenomenal.

According to the idealistic biologist, you may take any case of
reaction and describe it in this manner. There is a noumenal

428 IMPLICATIONS OF BIOLOGY

something which the noumenal organism selects from a noumenal
complex; and it is like that which we call oxygen, while the se-
lective process is like that which we describe as respiration. More-
over, the selected something is related to the complex from which
it is selected as oxygen is to air. Or again : when the roots of a
plant turn toward the center of the earth, they are noumenally
reacting to a stimulus which though in itself spaceless, is related
to the reaction as the centripetal pull of gravitation is related to
the downward turn of the roots. Or again : there is a noumenal
stimulus corresponding (mathematically) to light and a reaction
(adjustment and selection) corresponding to vision ; and the former
is of such a character that the noumenal organism must develop a
visioning organ (spaceless, of course) with a large number of struc-
tural differentiations which correspond to those differentiations
which we observe in the lens, the retina, the cornea, the aqueous
humor, the ciliary reflex, binocular accommodation, and many
others.

The advantages of this parallelistic hypothesis are conspicuous.
It avoids the fallacy of exclusive particularity,1 by admitting that
the pattern of the phenomenal series is, at least in some broader
feature, not "mere experience." And, by the same concession, it
is delivered from the ego-centric predicament ; for it accepts cer-
tain processes as not being constituted, in their pattern, by the
cognitive process (though constituted by it, with respect to their
elements2).

But there is another difficulty which it does not shun, and that
is the identity of indiscernibles. So far as I know, neither the
idealist nor his opponent has observed that the attempt to con-
strue the general situation of organism-and-environment as a
noumenal one results in the establishing of a one-to-one corre-
spondence between the relational patterns of the phenomenal and

1 Cf. 14, above.

8 This, be it noted, is the exact reverse of Kant's original hypothesis. Kant
regarded the patterns as created by the a priori synthesis, but the primordial elements
as somehow connected with the ' things in themselves.'

THE FALLACY OF INDISCERNIBLES 429

the noumenal systems ; and that this correspondence is made sig-
nificant only by making it not only a numerical but also a qualitative
and structural correspondence.

Take the adaptation of the flatfish. We would be setting up
a one-to-one correspondence between the phenomenal and the
noumenal orders, if we were to declare only that each discernible
peculiarity in the flatfish's adaptive reaction resulted from some
peculiarity in its noumenal environment. Schematically, this re-
lation would be : —

S 123456789

S' abcdefghi

in which S is the system of noumenal peculiarities, and S' the
system of the phenomenal. The extent and degree of corre-
spondence here is definable thus : for each element in S there is
one and only one corresponding element in S', and for each ele-
ment in S' there is one and only one element in S.

Now suppose that we make no further assumption. We have
not advanced by the minutest degree toward a characterization of
the biological situation. We have not indicated the specific char-
acter of the reaction of the 'psychoid' or 'ego' to the noumenal
environment. We have not even said enough to indicate that
there is a difference between types of reaction ; e.g. between dodg-
ing and eating, between breathing and perceiving, between cough-
ing and conjecturing. All that has been said is that in each reac-
tion each peculiarity of it meets some special peculiarity of the
objective situation. But we cannot pronounce the noumenal
operation 'vital' in the biological sense, unless the phases of its
activities are typically related to the environmental factors with
which they deal in the same general manner as the activities we
see animals performing are related to the environment we see them
dealing with. In short, so long as we assert merely that, to a
reflex involving, say, nine distinct acts there corresponds an environ-
mental complex having nine distinct efficient features, we have said
nothing that could give anybody the slightest reason for inferring that

430 IMPLICATIONS OF BIOLOGY

there is a 'psychoid,' or that there is an operation there like those
vital processes which we perceive, or that the vital processes differ
from one another (are specialized), or that they differ from inorganic
events.

To make sure, therefore, that the problems raised by reaction,
adaptation, evolution, mutation, heredity, and the like are genuine
problems, the biologist of conservative idealism is driven to af-
firm a likeness of pattern between the noumenal and the phenom-
enal differentiations. Schematically, this likeness may be thus
stated : —

S ^E: mnop

\R:w x y z

S' lE'im'n'o'

\R':w' x' y' z'

in which S is again the system of noumenal peculiarities, and
S' the system of the phenomenal ; and E and R the environment
and reaction phases of the noumenal system, and E' and R' those
of the phenomenal : and the small letters are the distinct features
within the phases after whose symbols they are written. Further-
more, there must be indicated the relation between E and R, within
each system; and this relation must correspond in kind to its
numerical correspondent in the other system. For instance, this
relation is sometimes that of selection: e.g. one or more of the
elements of E is picked out by R : w x y z. Or, again, the relation
may be that of avoiding an element of E by the system R:wxyz.
And so on. If this qualitative correspondence of relation is not
assumed, then surely, the 'psychoid' and the vital process can-
not be assumed (or inferred) to be anything more than products
of the cognitive act ; for the evidence that there is such a thing as
an organism, distinct from inorganic stuff, is all derived from just
these peculiar relations between spatially demarcated complexes
(living bodies) and their surroundings. There is a biological
problem, just because such bodies select some environmental ele-
ments, move toward others, perceive others, assimilate others, and

THE FALLACY OF INDISCERNIBLES 431

ignore others. And this biological problem is more than a prob-
lem of epistemology, in so far as these organic activities are more
than cognitive distinctions — are, in short, real behavior in the
ordinary sense of this word.

But how far has this concession led the idealistic biologist ? It
has led him to wipe out every namable difference between S and Sf.
This appears as soon as he attempts to interpret a specific case.
Take, once more, the imitative adaptation of the flatfish. The ele-
ments of the environment to which the creature reacts are, in terms
of percepts (our percepts or the fish's, as you will), some pebbles of
various sizes and shapes and colors, distributed in a certain irregu-
lar manner. The adjustment to the environment, again in phe-
nomenal terms, consists in the reproduction of this same pattern of
pebbles on the back of the flatfish (whether for protection or for
concealment from intended prey or for some other purpose is here
utterly irrelevant). To make the matter still more specific, let
us say that there is a sprinkling of round grayish pebbles over a
brownish sand ; the adaptation will then approximate this effect,
eventuating in round grayish spots on a brownish background
(within morphological limits of accuracy). The schematic ac-
count of this situation then is

S j fjf:

[R: w=£x=f=y=£z

S'. [E'-.m'^n'^o'^ p'

a ID/. /.£ 1+ I + I

In which =£ represents, in each case, a specific distinction between
the entities adjacently symbolized. Eg. in Er : m'=£nf, it repre-
sents the specific difference between a round, grayish pebble and
brownish sand. And in R' : w' =£ x', it represents the specific dif-
ference between a round grayish spot made by the flatfish's chro-
matophores and the brownish background made by adjacent
chromatophores. Please note that it does not indicate merely
that the terms it relates are logically different. It represents the

432 IMPLICATIONS OF BIOLOGY

concrete difference between them ; hence always specific characters
of them.

The biologist's difficulty now appears. He must say that the
relation between E and R is typically like that between Ef and
R', namely that of imitative adaptation to E by R. For, if it is
not imitative adaptation, then the relation of E' to R' indicates
nothing at all about the real situation. Secondly, m =£ n is typically
like m' =£ n' ; that is, the difference between a round grayish pebble
and brownish sand is, in some underlying respect, identical with the
difference between the two noumenal entities which set up the
noumenal reaction. That this must be the case, under the biol-
ogist's supposition, follows from the hypothesis (included in the
latter) that the phenomenal system is the result of the dynamic
relations between 'things in themselves' and the 'psychoids' or
'egos' ; and hence not a factor in the noumenal system. That is
to say, the flatfish does not adapt to the perceived differences be-
tween round grayish pebbles and brownish sand; it adapts to a
peculiarity in the noumenal environment. But, in order to use
what we see happening when the flatfish copies the color of the
sea-bottom, as evidence of a real process of adaptation, we are com-
pelled to say that the noumenal situation includes a difference hi
part identical with just this specific color difference ; and we are
so compelled, let me add, just because this specific difference be-
tween grayish and brownish is the one produced by the flatfish
reaction, and is moreover the only clue we have as to what is going
on.

Now, inasmuch as this is true of each distinguishable feature in
the reaction, and in every other reaction which we might scruti-
nize, it follows that the noumenal system consists of entities whose
relations to one another are, at least in some fundamental respect,
identical with the relations perceived by an observer ; and the nou-
menal reactions to those relations also are, in some fundamental re-
spect, identical with perceived reactions, and, finally, there is the
same fundamental partial identity between the relation of E to R
and that of E' to R', in each given instance. But if by hypothesis

THE FALLACY OF INDISCERNIBLES 433

we do not know the noumenal entities that are involved, we have no
means of distinguishing the noumenal order from the phenomenal,
other than these very relations. But these do not distinguish the
two systems ; on the contrary, so far as they testify, they identify the
pair, reduce it to a single system. The supposition that there is a
system beyond that which we perceive thus appears gratuitous.

The reader may suspect that this conclusion gains a specious
power through our having chosen as our illustration an extreme
case of imitative adaptation. To satisfy himself that this has not
happened, let him select for himself some other type of reaction ;
say one as far removed as possible from imitation. Let him ana-
lyze closely the reactions of a dog at which he throws stones. The
animal, far from imitating his action, concentrates attention upon
evading the missiles. The dog moves its body now to the right,
now to the left, now forward, and now backward, according to the
trajectory of the oncoming stone. If, now, these organic activ-
ities are real; if they do not derive their peculiarities from the
percipient of them, then the dog is trying to avoid being struck by
stones. If the latter be regarded as noumenal, as well as phenom-
enal, then they behave in a manner corresponding qualitatively
to the perceived motions of a perceived stone; and their action
includes characteristics in part (if not altogether) identical with
such perceived characteristics as 'two degrees to the right/
'fifteen degrees to the left,' 'upward,' 'downward,' 'parabolic
motion,' and so on. By what differentia, now, is such noumenal
activity to be held apart in our minds from ordinary perceived
motions ? There is none such. Noumenal and phenomenal are,
for us, indiscernibles, and therefore identical, provided we assume
the reality of organic reactions. There is no reason for inferring
that there is any 'situation' beyond that which we are directly
familiar with. There are only undiscerned items and complica-
tions in the situation we deal with. This last remark must be
guarded against an easy misinterpretation. It sounds like episte-
mological positivism, but is not. I am not saying that there
are only ' phenomena, ' as the positivist understands the term.

2s-

434 IMPLICATIONS OF BIOLOGY

Rather are there no phenomena, and no noumena, but only things,
events, conditions, circumstances all in a universe which no mind
has split into two realms. The positivist identifies phenomena with
appearances and then describes the latter as ' mental states,' with
the result that he limits human knowledge to ' the immediate ' or
to ' pure experiences ' and denies to it all ' things in themselves.'
Here is not the place to criticize this thesis; suffice it to state dog-
matically that it is rejected here.

IV

SOME DEFECTS IN MODERN THEORIES OF CONSCIOUSNESS

1. The Morphological Fallacy. — If, as is above maintained, there
is only one situation within which things happen; and if know-
ing is therefore itself a factor in that situation, it seems pretty
clear that no analysis of the inner peculiarities of cognition and
cognized entities (as such) will, of itself, explain the status of
cognition. Here we must expect just what we have always found
to hold true in physics, in politics, and in the affairs of daily life :
things reveal themselves most nakedly, not in their own nakedness
but rather when clothed in all their powers and activities and
relations to other things.

In the natural scientist's unqualified acceptance of this point of
view with reference to consciousness no less than to electricity or
tariffs, the difference between descriptive biology and speculative
philosophy appears. The biologist sets out with the situation
described in the opening of Part I ; accepting the presence of
organisms in an environment and of many types of action between
these, he undertakes to specify the relations which the constituents
of this enormously complex situation bear, one to another. For
him, the situation within which cognizing develops is enormously
greater than the situation which is cognized at any moment. If
he is interested at all in consciousness, he seeks out its relation to
the nervous system and to the many physical influences acting on
it and to the behavior which it sets up or qualifies. For him, there-

THE MORPHOLOGICAL FALLACY 435

fore, a description of whatever peculiarities may appear within
the cognitive field is useful by way of supplying symptoms that
may assist him in a diagnosis of that strange fever called knowing.
But it never crosses his mind to accept those structural peculiari-
ties as sure revelations of the mind's nature. That were no less
folly than to imagine that a few cross-sections of a nerve cell
would disclose the inmost secrets of nerve life and the utility of
nerves generally.

All that he knows cries out against such an error. The psychi-
cal functions generally do not terminate in perception or any other
type of cognition, but in various bodily readjustments ; and they
oddly seem to operate, with respect to any given matter, so as to
eliminate themselves after a while. As the aim of thinking is to
act without thought, so the goal of consciousness seems to appear
in the progressive dropping of percept elements and of members of
a reflective series, as the individual develops a habit of dealing
with the particular cognized matters. In short, they are con-
spicuously steps in a process, and are to be understood in the
light of it.

How unlike all this is the procedure of the philosopher who goes
on the assumption that epistemology is fundamental to all other
sciences! Consistently he believes that an analysis of the 'con-
tents of consciousness' will disclose the nature of consciousness;
and so, for several generations, his kind have been describing their
feelings and the flux of their immediate experiences, firm in their
conviction that introspection will carry them into the heart of
reality. From Locke to Bergson, so it has run on, and — as any
natural scientist might guess — from confusion to confusion.
Strange as it may sound in an introspectionist's ears, the under-
lying error of this method may be fairly described as the morpho-
logical fallacy, which is the mistake of identifying the structure of
an organic cross-section with the structure and function of the
organism as a whole. The stream of sensory characters does not
'look like' an ordinary cross-section, to be sure; it is a very
lively motion, full of lightning-like flashes and serpentine weavings

436 IMPLICATIONS OF BIOLOGY

of many, many things. But it is, for all that, only one phase in
the much more comprehensive organic process, and presumably
bears pretty much the same relation to this latter that the cross-
sectional motions in some one plane of a chemism bear to the total
chemism. Suppose one might peer into a constellation of cor-
puscles with a microscope of transcendent power. One would
there see, from any one given point of view, a vast tangle of mo-
tions, and yet discover nothing that would betray the peculiar
character of the chemism ; for all the motions that were significant
might occur in planes parallel to the observer's line of vision, and
their bearings might furthermore lie wholly beyond the micro-
scopic field. Now, just such a situation is given hi every act of
introspection. An infinity of elements which are absent count
heavily in the total organic process of which the moment's sensa-
tions are but a phase, and there is, within the introspected field,
absolutely nothing which indicates what those external factors
are or how the present factors relate to them. To overlook this
rather obvious fact and to construe the introspected field as the
legible sign post on the highroad to ultimate reality is to perpetrate
the grossest of abstractions.

There is not a little dash of irony in this circumstance. For
those thinkers who devote most attention to the structure and
activities of the primitive flux of consciousness are the very men
who find most fault with the ' abstractions ' and ' artificialities ' of
concepts (and even of percepts). They insist, with Bergson, that
the reality of things consists in the totality of their actions.1 But,
in the very next breath, they assert that, in the flux of immediate
experiences, there appear the deepest qualities of reality, such as
'unity,' 'homogeneity,' 'indivisibility/ 'pure activity.' Now,
I am not here urging a material argument against the genuine-
ness of these various predicates; I am only saying that, if one
does hold to the biological point of view, one cannot so exalt the
revealing power of the flux. To pronounce it the real and to label
the intellectual and other conditions to which it leads 'purely

1 Cf. Mati&re et M&moire, 25, etc.

THE PRAGMATIC POSITION 437

practical ' or ' abstract ' is possible, not through formal analysis, but
only through some presupposition about the very situation which is to
be analyzed. As was indicated in Part I, an analysis of any event
must cover, without prejudice, all the specific stages of the event.
The investigator cannot draw a line between any two conspicuously
different phases of it and call what lies on one side the real event
and what lies on the other its consequences or later circumstances.
He has the right to do something like this only after he has dis-
covered that the later phases do not depend exclusively upon the
former. If they do so depend, then former and latter are equally
real constituents of a single process. It is just this precaution,
however, which the introspectionist does not observe.

This accusation may be clearer, if put in the form of a figura-
tive query. Suppose that all light in the world around us were
very much mixed up ; all wave lengths running together, and all
fused in innumerable foci, except where they had passed through
certain rock crystals, liquids, and optic structures of animals.
Would a physicist be warranted in saying that the inmost nature
of light is revealed only in its interpenetrating mode, and that,
when spread out in a spectrum, it had suffered a serious diminu-
tion of its reality ? Certainly not; for that would be the morpho-
logical fallacy, the picking out of a single form from a variety of
forms belonging to one and the same entity-complex and attrib-
uting to it a metaphysical supremacy, without having proved that
all the other forms are genuinely (not verbally) dependent upon it.

2. The pragmatic point of view midway between the morphological
and the biological. — It was Dewey who broke the magic spell of
introspectionism and epistemology by insisting upon the philoso-
phers' return to the natural and illuminating biological point of
view. And he followed up this excellent advice with a reinterpre-
tation of the nature of reflective thinking, after the biologist's
manner. Wholesome as this reform has been, however, it has not
yet been thorough enough in its outworking. The direction of re-
form has been sensed more acutely than has the way of pursuing it.
All the pragmatists and the others who have thus far endeavored to

438 IMPLICATIONS OF BIOLOGY

describe mental operations biologically still fight with the weapons
of the enemy. They investigate the morphology of the cognitive
field. Thus Dewey accurately describes the broad situation within
which cognition arises and the general tendency of the cognitive
process ; but he deems of little relevancy the specific structure of the
material reacted to and of the material that gets into the cognitive
field. Furthermore, he describes the reaction chiefly in terms of the
ideas, meanings, and impulses which become conscious within the
reaction itself; and does not concern himself with its relation to
other reactions of the organism further than to point out that
these latter all make for a solution of a difficulty, a readjustment.
He shows how one idea leads to another, how a percept is tested
by another, how the ease of acting on a belief lends to the latter
a kind of value, and so on. He does not pretend to indicate the
precise relation of these movements to extracognitive conditions,
such as blood temperature, conduction currents, colloids, and all
the host of material factors which never figure discretely in the
natural operations of cognizing of the natural environment and
reacting to it.

Now, this analysis of the cognitive-reactive situation is neces-
sary, just as introspective analysis is. But it certainly does not
cover all the factors which shape and reveal what happens within
the cognitive field. It carries us much nearer to the biological
situation, but falls short of the latter in that it omits from con-
sideration the specific relation between cognizing and other organic
functions, as well as the specific relation between things ' in ' and
things ' out of ' consciousness. The inquiry ends with the indis-
putable observation that cognition facilitates practical conduct,
makes for success, assists the agent in adjusting himself to the
environment. Thus we are still left with the whole biological issue
on our hands: what kind of adjustment does cognition facilitate
which marks it off from breathing, perspiring, dodging, and all
the other bodily activities ? And hi what respect does a thing
cognized differ from the same thing uncognized, if at all ?

3. Analyses by new realists suffer chiefly from incompleteness.

ALEXANDER ON CONSCIOUSNESS 439

— It has been observed by several American realists that their
British brethren have not attacked the problem of consciousness.
They have successfully separated the problem of the entities from
the problem of cognizing, and they have dealt with the former
more thoroughly than any philosophers have, for many, many
years. But the latter problem is neglected, and so, when broached
at all, is answered vaguely. Thus it is when Alexander talks
of consciousness as a 'mental activity.' The difficulty of this
description is worth glancing at.

"Now what makes one thought-process different from another
is, I find, nothing but this difference of mental direction. . . .
I have no doubt that the thing called my consciousness exists,
and that it is mental activity. But it is not different in quality
as I am conscious of blue, or green, or the sun. . . . My con-
sciousness is one and the same thing working only in different
directions. ... I make no difference between . . . activity
itself, activity-consciousness, and the consciousness of ac-
tivity." »

Here something real is being described, and accurately. The
organism does aim at some things, and addresses itself to them in
such a manner that they fall upon the cognitive field. I should
not hesitate to say that the organism throws such selected
objects upon the cognitive field no less physically than it throws
them upon the retinas. Now, the activity of aiming and choosing
certainly neither constitutes nor is constituted by the things
aimed at or chosen. It is so no more here than in the case of
digestion or respiration. The lungs select oxygen, but do not
change their 'quality' as organs with every change of air.
They decide what shall fall within the respiratory field ; they shut
out noxious fumes that assail them and expel carbon dioxide. And
so too in what Alexander calls 'mental activity.'

Accepting the description, though, I must still wonder whether
it is not unenlightening to call consciousness 'mental activity.'
To define consciousness as that which works in different directions

1 Alexander, S., The Nature of Mental Activity, Pro. of Aristotelian Soc., 1907-
1908, 220, 225.

440 IMPLICATIONS OF BIOLOGY

is to put it in a class with every other organic activity. It does
not distinguish it from the antitoxic reactions of the blood. Cer-
tainly it is not poetry, but sober description, to say that the anti-
toxic chemisms set up by the diphtheria bacilli entering the blood
are directed at these bacilli. Certainly the bactericides get to
them and 'fix them/ and certainly they do nothing else. Is
their manufacture not a directed activity ? And is not the aiming
itself quite different from anything that a modern chemist would
pronounce a chemical action? Is it not that peculiar, little un-
derstood event which, in our ignorance, we cover with the blanket
term 'organic reaction'? And is not this term the more accu-
rate, inasmuch as we do not know what the aiming is ?

Well, if it is, then the same must be said of the term, 'mental
activity.' The aiming at things, in order to know them, cer-
tainly is not the same as knowing them. Neither is it the same as
thinking about them, for the sake of more information. It is no
more like cognizing than attending is like seeing. Indeed, I
should call it a case of attention, provided people would kindly di-
vest this term of its unfortunate psychological disguise. Attending
is a stretching out toward something ; it is not a feeling nor a know-
ing nor a thinking, any more than it is a digesting or a breathing.
It is rather the going to meet or to find some environmental char-
acter. It cannot be described in terms of the character sought,
for this varies from case to case, while the attending does not. It
would conduce to clarity, both in biology and psychology, if at-
tention were admitted to be a general organic attitude and not a
specialized function like cognizing. We might then speak of the
phagocytes as attending to bacteria without our falling into gro-
tesque panpsychism or idealism. But there is little hope of such a
clearing up. Psychologists will continue to bungle along, firm in
the conviction that whatever appears in connection with psychic
activities is necessarily psychic, just as the physicist thinks that
all occurrences in the material order are exclusively physical.

American realists have recognized that the problem of conscious-
ness is the problem of finding the differentia of the cognitive activ-

WOODBRIDGE'S THEORY 441

ity and that of the cognitive field. And so they have been trying
to discern the unique contribution which cognition makes to life
and to the entities which it groups peculiarly. At least three,
Woodbridge, Holt, and Montague, have named characteristics
which are truly enlightening; of these I would here mention only
the one Woodbridge brings out, as the others are discussed else-
where in this volume.

In a noteworthy essay x Woodbridge sets forth that conscious-
ness involves a type of relation between entities which obtains
nowhere else; namely, the relation of implying one another. That
the relation referred to is really present, and that it is significant,
I have no doubt. But it stands in need of extensive explanation,
even as the relations pointed out by Holt and Montague do. We
must know more about it, before it illuminates us. For instance,
is it what logicians mean by implication ? Is it established by a
psychic reaction, or revealed by it, or is it the reaction itself ?
How is it connected, if at all, with what the psychologist calls
' simple association ' ? Woodbridge raises these queries in his
readers' minds when he says that bread implies nourishment, and
that implication is deeper than the inferential relation between
propositions. But I do not believe the queries can be answered,
until the whole biological situation has been resurveyed, and the
relations between environment and organism stated more precisely.
This, if I mistake not, is Woodbridge's own belief. And the same
restrictive judgment must be passed upon every other realistic
hypothesis. Consciousness is not described in terms of the total
organic situation out of which it develops and of which it is an in-
tegral phase. One hypothesis deals only with the relation of per-
cept to object, in terms of these alone. Another analyzes simply
the structure of logical entities. A third considers merely the
manipulating of percepts in an organic readjustment. A fourth
focuses upon the psychophysical processes. Is it not possible,
therefore, that all may be, on the whole, equally valid just because
each attends to something different ? And may their harmony not

1 The Nature of Consciousness., /. of Phil., Psychol., etc., 2, 119.

442 IMPLICATIONS OF BIOLOGY

appear, as soon as consciousness is investigated, as a feature of a
big situation which involves not only feelings and flux and think-
ing, but also an organism, — blood and sinew and nerves and im-
pulses and appetites, — and also physical things, — electricity,
light, matter ? At all events, the bare possibility of this is a suffi-
cient warrant for attempting to realize it.

4. The biological point of view. — It must now be clear that no
investigation of consciousness is truly biological unless it first de-
scribes consciousness in terms of the three basic factors : stimulus,
reagent, and organic instrument (structure-function). The prob-
lem may be stated in three corresponding forms.

First of all, consciousness involves a specific environment. This
we must discover and describe. The cognitive function copes with
some peculiarity of the environment which no other function
handles. As the stomach does something which no other organ
does, so too the central nervous system, and, more specifically, the
cognitive reagent.

Secondly, consciousness involves a directed activity. That is, it
is not simply set going by some stimulus; but, when set going,
it heads somewhere. Perhaps it is not purposive ; but assuredly
it has a direction just as truly as each individual case of magnetic
pull or of gravitation has a direction. (Whether its direction is
purely spatial, or spatio-temporal, or of some other order, is a
special question.) It 'gets somewhere.' It accomplishes some-
thing. But what ? This also is to be discovered and described.

Finally, consciousness involves the operation of an organic
structure, and, according to all that we know about structures
and functions, the structure somehow modifies that which, in
affecting it, eventuates in cognition. Now are any peculiarities of
consciousness due to the structure ? If not, what does the struc-
ture accomplish? Once more we must discover and describe.

From the biological point of view, these three problems are ir-
reducible and equally important. Let any one of them be ignored,
and the presumptions are heavily against anybody's answering
unequivocally the two central questions : In what respect does

BIOLOGICAL STATUS OF CONSCIOUSNESS 443

an entity 'in' consciousness differ from an entity 'out of con-
sciousness ? And what does consciousness accomplish ?

THE BIOLOGICAL STATUS OF CONSCIOUSNESS

THE reader will kindly note that this section proposes to discuss
only the status of consciousness in the realm of life. It will offer
no completed hypothesis, but will only describe the conditions
under which consciousness arises. The most that can be claimed
of this description is that it narrows very closely the range of pos-
sible interpretations.

1. The general structure of the environment. — It is a singular fact
that no modern philosopher nor speculative biologist has formally
analyzed the broader features of the world in which individual
organism exists. No fact shows so strikingly as this one how the
anthropocentric world-view has dominated contemporary thought.
Pick up any textbook on psychology, and you will find extensive
descriptions of sensory qualities and nervous structures, but of the
external field of stimulating entities not a word, save such plati-
tudes as these : " Luminous bodies affect the retina with ether
waves." — "Pulses of air are heard as sounds," etc. To be sure,
the experimental psychologist is constantly driven to analyze
special physical features, in connection with his special problem.
He informs himself about acids, salts, and alkaloids, if studying
taste and smell ; and about light, when investigating vision. But
always he attends to their minute peculiarities, never to their
common and collective properties and conditions of activity, with
respect to organisms. By such properties and conditions I mean
such as either shoot through all or most of the physical entities with
which organisms deal, or else are involved in the coexistence and
coactivity of those entities.

Three properties or conditions of this sort are (a) space, (6) time,
and (c) the causal relation. There may be many others, but these

444 IMPLICATIONS OF BIOLOGY

are obvious ; and so, too, is their biological importance. Every
living creature finds itself in a world full of things distant from it
and from one another in space and in time. Some of these things
it seeks, others it shuns; and the precise relation of particular
things to its body in space and in time is a life-or-death matter.
Be one's metaphysical theory about space and time what it will,
in any case one must grant that potatoes in the lungs instead of in
the stomach would affect a man's welfare more substantially than
the loftiest philosophy ever written; and that the history of
species and of civilization might be largely written in terms of
'before' and 'after.' The individuals which have survived and
shape empires and ideals are those whose ancestors have been
reaching conclusions, having inspirations, carrying out plans of
action, winning help, reaching places of safety a little sooner than
somebody else. As for the value of mastery over causes, this has
been dilated upon since Roger Bacon ; and has never been doubted
by the ordinary man.

Now, is it not likely that, just as the special differentiations of or-
ganic form and function are connected with special characters of ele-
ments in the environment, so the more inclusive are connected with the
commoner features of the environment as a whole? And, inasmuch
as every organ save the central nervous system deals specifically with
some special material which encounters it from moment to moment,
and inasmuch as the central nervous system specifically reacts to all
kinds of materials, even those in remote spaces and remote times ; is
it not probable that the deepest utility as well as the deepest peculiarities
of cognition can be understood only through an analysis of the deepest
peculiarities of the whole space-and-time order of nature?

These questions have never before been raised, at least not with
the earnestness which they deserve. Consider, for instance, the
manner in which philosophers have scrutinized space. With the
exception of the Aristotelians and the materialists, who have at
least accepted the full reality of space, even though they have not
reckoned sufficiently with it, philosophers have handled this mode
of the environment in one or more of the following three ways:

ANALYSIS OF ENVIRONMENT 445

(a) they have depicted its supposed paradoxes — so from Zeno to
Bergson ; or (6) they have regarded it as either an idea or a mere
form of apprehension and have traced its genesis and the arrange-
ment of its psychic constituents — thus from Locke through Kant
to modern psychology and metaphysic; or finally (c) they have
joined with the mathematicians and have analyzed the pure form
of externality, as it is conceived in geometry, but have not regarded
their results as having any bearing upon the biological situation.
In this persuasion Mr. Russell finds himself when he declares that
protective geometry is 'wholly a priori,' taking nothing from expe-
rience and having ' like Arithmetic, a creature of the pure intellect
for its object.' *

Of these three attitudes, only the last bears off in the right direc-
tion. Certainly geometers have analyzed space with exceeding
rigor. But, when they have not believed their subject matter
purely intellectual, they have not been interested sufficiently in
its possible biological bearings to describe the environment in
terms of it. Indeed, most of them are so sure that geometry is
' a creature of the pure intellect ' that they look with pity upon the
thick-witted realist who, doubting that the pure intellect has any
creatures of its own, seeks the facts of geometry in the world-order.

2. The environment as a space complex. — Suppose we look at the
spatial bearings of the world in which animals have to seek their
food and drink and shelter and mates. And let us try to find just
those relations which obviously 'make a difference' or 'generate
a problem ' for the creature which has to inhabit the space. First
to be noticed are these :

(a) Objects important to life are mostly masses of matter giving
off various types of energy. (The term 'energy' is here used in
its colorless meaning, and does not imply any physical hypothesis.)
Thus, all articles of food and drink ; each is located narrowly at
some region in space. So, too, are most places of shelter and refuge ;
so too are other animals, foes, prey, clan mates, and creatures of the

1 Foundations of Geometry, Cambridge, 1897, 118. Perhaps Mr. Russell would
qualify this statement to-day.

446 IMPLICATIONS OF BIOLOGY

opposite sex. So, too, is the sun, on whose light life depends. So
too are most objects which assist an animal in finding its bearings
and making its way up and down the earth. Indeed, there seems
to be only one universal commodity which is not localized but
diffuse, and that is air.

(6) Most of these objects are centers of influences extending
far beyond the masses of matter which men commonly call 'the
things themselves.' For instance, the sun sends forth ether
vibrations ; objects reflect these vibrations according to their own
chemical structure and so give off peculiar configurations of light
(color-form complexes). Fire emits heat, and mountain lakes
make their coolness felt at a distance. Flowers send forth
odors, and so do most animals.

(c) In most cases, these influences go out in all directions of space.
The objects are, in the wider sense of the term, radioactive.

(d) An animal is related to practically all the elements of its
environment through these radial lines of energy. The only
elements to which it is not so related are those which it takes
bodily into its organism. And even with these it is radially
related up to the moment of receiving them.

(e) In all observed instances, the character of any part of a
given energy radius is a (continuous) function of the distance from
the energy center. (It may at the same time be a function of some
other variable too.) Heat, light, gravitation, odor, and whatever
other qualities are radiated vary from point to point along each
energy radius in some inverse ratio to the distance of the particu-
lar point from the energy center.

(/) Each energy radius is, in its geometrical form, a real line.
That is to say, with respect to its spatial features alone, it is a con-
tinuous series, viz. one which satisfies Dedekind's postulate, the
postulate of density, and Huntington's postulate of linearity.1
Unlike a pure geometrical line, however, it is infinite in only one
direction at most, and possibly finite in both directions. Strictly

1 Cf . Huntington, E. V., The Continuum as a Type of Order, Annals of Math.,
Second Series, 6, 15.

ANALYSIS OF ENVIRONMENT 447

speaking, then, it is a segment of a geometrical line. But, inas-
much as the environment of any given animal lies wholly within
such segments and the problems generated by space derive from
the linearity of the segments (as will be shown), this circumstance
makes no material difference in our present inquiry.

(gr) Each energy radius is not only a spatial line, in the above
sense, but also a continuum of physical qualities other than spatial.
Superimposed, as it were, upon the point series between my eye
and the sun is a series of ether conditions. While this superim-
posed continuum may not be of the same structure as the point
system, it is always some kind of a continuum or a series of con-
tinua which pass from one to the other through 'critical points'
(which involve some kind of qualitative change).

(h) Energy radii from indefinitely many energy centers inter-
sect, sometimes without reciprocal modification, sometimes with
it. Rays of light intersecting in a focus are not altered by that mere
act. But they are when passing through glass, which refracts,
reflects, and absorbs certain of their characters.

Two types of problem are generated by this circumstance : first,
the locating of an energy center; and secondly, discovering the
specific characters of an energy center, as distinct from those de-
rived from intersections with radii from other centers.

These two problems just named are, strictly speaking, not
generated by the environmental character exclusively ; but rather
by it in combination with the specific needs of the animal. Though
the latter do not belong in our present enumeration, I shall cite
them because they derive partly from an important quality of
physical qualities. The animal sometimes needs to get an object
'in itself (that is, the energy center) and sometimes it needs only
a distant influence of such an object. Consider a man and a
hearth fire, on a bitter winter's night. So long as he wishes only
to warm his hands or his back, his problem is not to get the burn-
ing wood as he might get food, but rather to hold just the most
favorable distance from it. If, however, a spark shoots out on to
a rug, his problem is to pick it up or stamp it out ; and now it is

448 IMPLICATIONS OF BIOLOGY

not an influence but an energy center itself to which he must adjust
himself. Likewise with respect to the other than spatial charac-
ters of the fire. Sometimes he must know whether it is wood or
coal that is burning.

(i) No single point on an energy radius determines, contains,
or indicates its own distance from its own energy center. For
distance between points is an external relation in both the purely
logical and the geometrical sense of the term. The points, as such,
do not constitute the distance, nor are they constituted by it ; and
the distance is their externality, not their position.

(j) No single point on an energy radius determines, contains,
or indicates its own direction from the energy center. This may
be simply stated in the usual language of geometry. Two points
are required to determine a line; for the line, as Mr. Russell
phrases it, "may be regarded as a relation of the two points, or an
adjective of the system formed by both together." * Hence, if
only one point is given, all lines passing through it are qualita-
tively indistinguishable.

Observations (i) and (j) do not depend upon geometrical demon-
stration for their accuracy. Anybody may verify them empiri-
cally. In his Essay toward a New Theory of Vision, Berkeley
said : " Distance being a line directed endwise to the eye, it projects
only one point on the fund of the eye — which point remains in-
variably the same, whether the distance be longer or shorter." 2
Though the definition of distance the Bishop here gives is inaccu-
rate, the gist of his remark is correct. Indeed, it holds good not
only of single points but of many (though not all) point-complexes.
Thus, taken in isolation, a perceived line may be a short one near
at hand or a longer one farther away or a prodigious stretch at a
great distance. Only certain combinations of real points in cer-
tain relations determine their energy centers' directions and dis-
tances. Hence,

1 Foundations of Geometry, 120.

1 His inference from this, namely, that distance could not be seen, was unwar-
ranted; and for reasons which will soon be indicated.

THE PROJECTIVE RELATION 449

(K) Direction and distance are functions (mathematical) of certain
types of complexes, at least insofar as the kind of system above de-
scribed is concerned. So stated, the fact will strike many persons
as paradoxical. And yet it is nothing but the exemplification,
in the physical world, of the facts on which projective geometry
is founded. These facts appear, in brief, to be identical with those
which describe the environment in its dynamic relation to organ-
isms. Indeed, the latter relation can be stated wholly in terms
of projective geometry; and when so stated, it reveals just that
peculiar state of affairs which involves consciousness.

(/) The projective relation obtains in systems of externality which
do not contain metrical values. — Geometers have recognized that
quantity is logically posterior to quality, at least in spatial rela-
tions. Qualitative identity is presupposed by every quantitative
identity. "Hence," to quote Russell, "all figures whose differ-
ences can be exhaustively described by quantity . . . must have
an identity of quality, and this must be recognizable without an
appeal to quantity. It follows that, by defining the word 'quality'
in geometrical matters, we shall discover what sets of figures are
projectively indiscernible." Now it appears that the pure qual-
ities of space are (a) points and (6) their differentiating relation,
namely, the line. The relation between two points is the straight
on which they lie. "This," Russell says, "gives that identity of
quality for all pairs of points on the same straight line which is
required. ... If only two points are given, they cannot, without
the use of quantity, be distinguished from any two other points
on the same straight line ; for the qualitative relation between any
two such points is the same as for the original pair, and only by a
difference of relation can points be distinguished from one another." 2
Conversely, too, a straight is only the relation between two of its

1 1 do not attempt to demonstrate here that projective geometry is non-quanti-
tative. I shall only indicate it and leave the reader to such treatises as Veblen,
O., and Young, J. W., A Set of Assumptions for Projective Geometry, Amer. J.
of Math., 1908, 347, etc. ; Russell, B., loc. cit., Ill, Section A. ; etc. The latter is
recommended to the layman.

*Loc. cit., 130.
20

450 IMPLICATIONS OF BIOLOGY

points and is distinguished from another only by them. Hence,
given only one point, any pair of straights is qualitatively indis-
tinguishable from any other.

I wish to emphasize at once that there is nothing mysterious
in this fact, and also that it is not peculiar to a recondite branch of
geometry. It is true of all kinds of entities. A character which
defines an entity as a member of a certain class cannot distinguish
it within that same class. Suppose we learned just what it is that
makes a horse a horse; that is, the complex of characters which
distinguishes the equine quality from the asinine, the bovine, etc.1
We should not possess the information necessary to distinguish
a Percheron from a Clydesdale, or a Morgan from a French Coach.
Qualitatively — which here means 'with respect to equinity' — all
horses are, qua horses, indistinguishable. A platitude of elemen-
tary logic in this context, our observation becomes enormously
difficult and obscure to many thinkers when it refers to space
classes. They cannot perceive that what makes a position a posi-
tion and a line a line is not the same as that which makes one point
ten feet from another, or to the left of another. And, missing this,
they are simply bewildered by our earlier statement that direc-
tions and distances (spatial quantities) are functions of point-line
complexes, not of point-quality and linearity.

(m) In both pure geometrical projection and physical projection
there are certain complex relations which are constant qualitatively,
regardless of the distance or direction of the related elements
from the projection center. In pure geometry the fundamental
relation of this type is the anharmonic ratio. If through any
four points in a straight four straights be passed meeting in
any point, and if another straight meet these same four, then the
four new points of intersection have the same anharmonic ratio
as the original four points.2 Or conversely ; all lines cutting any

1 For the present argument it is irrelevant that organic types are not sharp,
stable things, as they were once supposed.

* For a proof that this ratio does not involve distances or angles, cf . Russell, loc.
cit., 122, 125.

THE PROJECTIVE RELATION 451

four lines that meet in one point have therein the same anharmonic
ratio. In physical projections wholly different qualities of radial
systems exhibit a logically similar constancy. Thus, a pencil of
light rays from a given luminous point may be cut at any point by
a curve whose generating center is the luminous point; and the
sum of heat or light energy at all such cuts is a constant.

How many such projective constants there are, nobody knows ;
but geometry, physics, and psychology bring forward facts indi-
cating that the variety of types is exceedingly great. When we
come to discuss the peculiarities of the cognitive field, we shall
encounter some striking kinds. Comparing them, we discern a
highly significant fact which, on closer inspection, turns out to be
no accidental feature of the projected matter but an intrinsic
feature of the projective relation. It is this:

(n) The relations which are constant and determinate in any given
projection are, in the logical sense, transverse to the lines of projec-
tion.— To make this clear, let us look at the generic properties of
projection as exemplified in the very simple specimen; namely,
the projection of points upon a plane. As its properties have been
thoroughly worked out, we may turn to it without fear of confu-
sion. I shall employ special names as generic terms which, when
applied to cases of projection of higher order, will not be colored
with the specific suggestions of geometry.

(i) If, from a fixed point S, lines be drawn to different points,
A,B,C, • • and if these lines are cut by a plane in points A', Bf, C',
. . ., the latter are termed the projects of the given points upon
the plane.

(ii) I shall designate as 'the projection system' the total com-
plex consisting of the three parts : (a) the fixed point S, (6) the
collection of points, A, B, C, - • - ; and (c) the plane of the points
A', B', C', • • • I shall designate the collection of points, A, B, C,

• • - , as the projected complex; and the collection of points,
A', Bf, Cf, • • • , the project-complex ; and the plane of A', B', C',

• • • as the projection field. The point S I shall name the pro-
jectorial referent.

452 IMPLICATIONS OF BIOLOGY

(Hi) The term 'order' shall designate the order of dimen-
sionality of a complex. Thus, Euclidean space is of the third
order ; the space-tune system is of the fourth order (at least) ;
and so on.

These are, I believe, all the fundamental definitions needed for
an understanding of the generic properties of projection (as dis-
tinct from specifically spatial projection). What, now, are these
generic properties ? They are four in number :

(i) If the projected complex is of the nth order, then the pro-
jection field may be of the (n + a)th order.

This follows directly from the fact that the projection field is
external to the projected complex and also to the projectorial
referent. This externality may constitute a dimension over and
above the dimensions of either the projected complex or of the
projectorial referent. This is readily seen in the specific case of
point projections upon a plane.

This projection system is of the third order, for the plane (Ar, B',
C' D') is external to both S and the points A, B, C, D. Hence the
line AS and that plane form a tri-dimensional complex.

(ii) In this construction the projection field necessarily is (a)
transverse to the relation between the projected complex and the
projectorial referent, and (6) external to both of these. It is clearly
seen in the specific case above given. The elements of the pro-
jected complex there are points, but the relations between the ele-
ments of the project complex are lines. Secondly, the relation
between any given element of the projected complex and the pro-
jectorial referent is a line, but a line of a plane other than the
projection field and hence of another dimension than any in the
latter. It is also worth while to note, thirdly, that the relations
between the elements (points) of the projected complex are lines
of other planes than that of the projection field.

(in) The relations between the elements of the project-complex
are of a different dimension from (a) the relations within the ele-
ments of the projected complex, and (6) the relations between the
projected complex and the projectorial referent.

THE LIFE SITUATION PROJECTIVE 453

(iv) The relations of elements to one another within the pro-
ject-complex are a function of (a) the character of the projection
field, (6) the relation of the projectorial referent to this field, (c)
the relation of the projectorial referent to the projected complex,
and (d) the relation of the elements of the projected complex to
one another. Thus, in the figure above, the relation of A' to B'
depends upon the structure of the projection field — e.g. whether
it is a plane or a certain curved surface or some other type. No
less does it depend upon the distance and direction of S from the
projection field and from the projected complex, and upon the
relation of A to B.

These properties of a projection system are, I believe, strictly
generic. They are to be found wherever there is any kind of a
projective field typically related to a projectorial referent and to a
projected complex. And it matters not what the specific dimen-
sions of the complex are ; whether they be the familiar space and
time dimensions, or some other kinds, such as color dimensions.
The properties seem to be those of 'pure projection,' just as com-
mutativity is a property of pure number and free mobility a prop-
erty of pure space. Now, if this is a correct opinion, it is natural
to wonder whether some specimens of projection are to be found
outside of pure geometry. It seems to me that a remarkable one
is given in the organic situation of which consciousness is a
phase.

3. The three factors of the biological situation correspond to the
three factors of the projection system. — The reagent is the pro-
jectorial referent, the environment is the projected complex, and
the cognitive field is the projection field. The difficulty which
most of us find in comprehending this is due to two circumstances :
first, to the naive view of the projective relation, as a throwing of
simulacra on a screen ; and secondly, to the almost inevitable tend-
ency to conceive the true projective relation as occurring within
tri-dimensional space, instead of being a relation between the
spatio-temporal order and a higher one. The first of these diffi-
culties I have tried to clear up in earlier passages ; the second will

454 IMPLICATIONS OF BIOLOGY

be at least mitigated, I trust, by the following account of the re-
agent and its projection field.

(a) The reagent is distinct from (i) that to which it responds,
and (ii) its own response. Earlier in this essay it was shown that
reactions of the ordinary kind cannot be adequately stated in
terms of their stimuli alone, but that there must be assumed some
entity (not necessarily psychical) whose attitude or reference to
the stimuli defines the reactive relation. It is opportune now to
show this same fact in connection with consciousness.

The difference between the stimulus pattern and that of the
reaction is hardly worth dilating upon, so evident is it. If I see a
box three feet long and two feet deep, I do not become a box three
feet long and two feet deep. If I hear that a man fell dead in
Broadway yesterday, I do not fall dead in Broadway yesterday.
In short the reaction is not a simple duplicate of the physical cir-
cumstances inducing it. It bears off in its own peculiar direction.
The clearest case of this is the time reaction. Past and present
stimuli set up reactions which are regulated by future conditions ;
and the tune pattern of the reactions contains future elements,
which, of course, are not in the physical stimulus at all.

Before we attempt to explain how this pattern difference is
brought about, let us observe the second peculiarity, namely the
externality of whatever does establish the before-mentioned dif-
ference. Here again, I am stating only a commonplace in language
stripped of presuppositions. I refer to the well known fact that
what a man does, as a result of cognizing some situation is not
deducible from the internal structure of that situation alone ; and
still less from the internal structure of the resulting act alone. His
conduct is determined primarily by something which philosophers
call by various names, such as 'impulse,' 'appetite,' 'vital force,'
'psychoid,' or 'ego.' x

1 These last three designations are so steeped in bad metaphysics that their use
leads almost inevitably to a misconception of the whole situation. I cite them here,
simply to indicate that those thinkers who employ them are referring to a real factor
in the life process.

THE PROJECTIVE SITUATION 455

Sound a horn in the presence of ten men. One hearer will clap
his hands to his ears and cry: "Stop that din !" A second will
fall into a revery of pleasant melancholy. A third will dance. A
fourth will stand at attention, like a soldier. Now, these differ-
ent behaviors are not deductions from that which is cognized.
They are chiefly attitudes toward it. The attitude may be taken,
for aught I know, by some group of cortical cells, or by the blood,
or by a blithesome archangel ; but it is taken, and both the agent
and the act are not parts of the situation toward which the atti-
tude is taken. Likewise of the ensuing conduct. In the mere
sound of a horn there is no potentiality of clapping a pair of hands
to one's ears, nor of revery, nor of dancing, nor of unconscious-
ness. Nor do these powers lurk in the consciousness of the sound.
To adopt the overworked and inadequate phrase of the biologist,
they are all activities which the cognized sound releases, as the drop
of a trigger releases the explosive forces of gunpowder. More
precisely, the cognized sound is the critical element in a process
involving much more than the element and specifically determined,
though not incited, by other factors than that element.

(6) The relations to which the reagent responds through the
help of consciousness are relations among spatio-temporal entities,
but they are not spatio-temporal relations, in the strict adjectival
sense. That is, they are not distances nor directions nor magni-
tudes nor durations. These latter and other similar kinds are,
in the cognitive field, terms among which relations of wholly
different orders obtain. What these specifically cognitive rela-
tions are is a question too extensive for the pages alloted me ; but
I should like to say at least that Woodbridge has long since pointed
out the most important class, namely, implications. These cer-
tainly are not spatio-temporal relations, in the proper sense of
being developed or present as efficiences in physical and chemical
processes ; and yet they are not created by being known, they are
not mental devices, but real relations between real entities. Or, to
put it in a phrasing that escapes the suspicion of subjectivism:
the stimuli, as physico-chemical processes, are really implicated

456 IMPLICATIONS OF BIOLOGY

with many other events (entities) in the universe, but this implica-
tion is not present within the chemisms any more than the direction
or distance of a point from another is present within the point. The
implication is a relation between entities, not internal to any one
entity; hence the relation appears only when the complex of in-
volved entities does, just as linearity, straightness, curvature, etc.,
appear only when a complex of points does.

What, now, is the complex in which implicative relations appear ?
The answer to this question designates the unique status of con-
sciousness in the world. The complex embraces the entire physi-
cal universe, past, present, and future, together with all types of
relations.

(c) The reagent is so related to the projection field and to the pro-
jected complex, that, in some dimensions, the projection field lies be-
tween it and the projected complex. This is only an exact way of
stating that our organic adjustment to physical things is 'through'
the field of consciousness.

The projection field, or the 'field of consciousness/ is, in the
strict logical sense of the adjective, transverse to the objects pro-
jected upon it. Transverse means different with respect to a given
dimension. This is precisely the relation of the cognitive field
with respect to the elements in it. The latter are spatio-temporal,
but the relations between them in the cognitive field are not spatio-
temporal. None of them can be reduced to length, breadth, thick-
ness, duration, or to any complex of these. For the very same
reason, then, that length is a dimension different from breadth,
a cognitive relation is of a dimension different from any spatio-
temporal one. And the field of cognitive relations is in that differ-
ent dimension. It is transverse to the four spatio-temporal
dimensions and is therefore of the (4 + a)th order. If we agree
to define the physical world as the spatio-temporal system exclu-
sively, then consciousness is not physical.1 But this does not imply

1 It may be well to add that I should reject this definition. I do not see any
strong evidence for supposing that physical entities are mere complexes of extension
and duration.

THE PROJECTIVE SITUATION 457

that the objects of consciousness are not physical. Nor does it
even imply that cognitive relations are not relations between
physical things.

The broader features of the whole process may now be summar-
ized : Consciousness enables the organism to adjust itself in various
manners to entities external to it in space and in time at the moment
of the specific conscious state. I do not say that consciousness
consists in the specific adjustment; this is predominantly impulse
and motor performances. But consciousness is the crucial ad-
vance toward this adjustment. It makes possible my regulating
my behavior, here and now, to physical objects which have ceased
to exist, to others which have not yet come into existence, and to
existent objects which are not affecting me in space at the present
instant. An incidental feature of this capacity is the cognizing
of purely imaginary and impossible objects and of behaving toward
them. Sometimes I take them seriously, and then, so far as their
immediate efficiency is a criterion, they are just as real as physical
things; and I am victimized by my own consciousness. But
usually I am only curious about them, or I laugh at them, as I do
at the Walrus and the Carpenter.

What must be involved in this procedure ? Note first that the
environment which enters the cognitive field is four-dimensional.
The organism which adjusts to it variously is also four-dimen-
sional. I am genuinely conscious of all three dimensions of space,
and also of time ; no one of these dimensions goes lost in entering
the cognitive field. And I react in these four dimensions too;
my muscles not only move in the three dimensions of space, but
they defer or accelerate their motions in time. The transition
from the cognized complex to adjustment involves a projection
of the former upon a field of the (4 + a)th order. The specific
relations within the projection are functions of the characters of
the four-dimensional complex that is projected. But, although
they are functions of these, the specific relations within the projec-
tion are of a different order. Precisely the same holds of the pro-
section characters and the organic adjustments they induce.

458 IMPLICATIONS OF BIOLOGY

In less exact language the function of consciousness, when
viewed with respect to its direction, is to make efficient in the four-
dimensional organism those types of relations which are peculiar
to fields of the (4 + a)th order.

Or again, in language grievously tainted with old philosophical
errors, consciousness is that efficiency whereby we regulate our
conduct according to principles which are 'above' space and time.
This way of stating the case is not illuminating, in spite of its
familiarity. It leads us into all the stock puzzles of epistemology,
just because it describes only the result of consciousness and not
the specific relation of the involved factors to one another. It is
therefore to be discarded in favor of the less simple but more
descriptive account above given in outline.

4. If the biological situation constitutes a projection complex,
as has been maintained, then all those events which, in psychology,
are called illusions and errors are not products of consciousness, but
they are peculiar and inevitable characteristics of the whole pro-
jection system. They stand in the same relation to the 'total
situation' that a projected point does to a purely geometrical
projection system. Space does not permit an adequate exposi-
tion of this matter here ; but I cannot dismiss it without a brief
statement. For, as Holt and Montague have shown, the crucial
problem for the new realism, as for every other theory of cogni-
tion, is the problem of error (in all forms) . And the acutest critics
of the new realism urge that its fatal flaw is its acceptance of the
full 'objective' nature of illusions and error and its simultaneous
refusal to put illusory objects, together with all their colors, shapes,
and behaviors, identically in the very space and time in which
they immediately appear. If the charge is true, it is deadly. But
I believe it owes its strength wholly to a misunderstanding of the
relational character of the "organic situation" in which conscious-
ness develops, and of the entities which enter into that situation.
This situation contains a host of entities which are protectively indis-
cernible and which therefore must possess, in any given projection
field, multiple values. These entities are not merely construed

PROJECTIVE INDISCERNIBLES 459

erroneously ; they are genuinely indeterminate with reference to
our spatio-temporal systems of which they are true parts or
phases.

In any projective system, any given project-complex is the pro-
jection of an infinite number of real or merely possible projected
complexes. Thus, if from the point S the lines SP, SQ, and SR be
drawn cutting another line in A', B', and C" ; then A' is the projec-
tion of each point in the line SP, B' of each in SQ, C' of each in SR,
Given only S and the projection plane and the values A', B', C",
therefore, each of the latter is the true projection of an indefinite
system of genuine indiscernibles and has therefore multiple value.

In order to determine which the real projected complex is at a
given instant, we must test it with other systems of points,
M , N, 0, which must be varied, while S, the projection plane and
A', B', C", are constant. If, then, A', B', C', does not vary with
variations of M, N, 0, the latter complex is not the real projected
one. And so on. Please note that this elimination involves a
series of operations, much more than mere projecting. It in-
cludes a directive activity, comparison of project complexes, and
selection among these.

These projective indiscernibles are precisely the relations which
we find, much more complicated, in the much higher order of pro-
jection on the cognitive field. Substitute for points enormously
intricate entities of the fourth order or higher. Substitute for
the projection plane a field of unknown higher order. Substitute
for S a projectorial referent which, instead of being of zero dimen-
sionality, may be of a high order. And you then have (i) a center
to which a distant complex is related ; (ii) this relation a function
of space and time and perhaps other dimensions; (Hi) and the
projective relations among the projects transverse to the dimen-
sions in which the projected complex is related to the organic cen-
ter. For example : When I think that yesterday's sunset means
rain this afternoon, I am external to the sunset and the rain in
space and in time ; my relation to them is a function of space and
time (and probably more variables), but what I think about the

460 IMPLICATIONS OF BIOLOGY

sunset is not, in itself, a complex of spatio-temporal relations, but
rather a complex of suggestions, implications, discriminations, and
the like, some of whose terms may be spatio-temporal, however.

Now is it not a familiar experience that (i) no single complex in
the cognitive field ever has a term that is not of multiple value, and
(ii) conversely, with reference to a single cognitive situation, there
is an indefinite number of projective indiscernibles for each project
element ? Thus : a perceived line, at a given single instant, suggests
or implies a line of the length m at the distance n and in the direc-
tion a; but, so long as no other cognitive situation is introduced and
the behavior of the percept in the new situation is not compared
with its behavior in the former, the implied line may not be the
one which is actually projected at the instant. A longer line at a
greater distance may be perceived, and in perception may be at
the instant genuinely identical with the implied line ; which is to
say merely what everybody knows directly, viz. that many things
different in other contexts are identical in one perception, i.e. are
there indiscernibles, having all one perceptual projection, pre-
cisely as all the points in the line SP have one common projection
in the given plane, under the given circumstances.

And is it not also a matter of common observation that the same
holds of concepts, no less than of percepts ? It is not an accidental
feature of visual space, but is involved wherever a purely projec-
tive (non-metrical) relation is. Thus, the specific, momentary
'idea' I entertain is the projection of an indesignate multitude
of entities each one of which implies and is implied by the very
same other entities when projected into the particular cognitive
field. I think, in a given cognitive situation, of 'a brown mare
that is gentle and saddle-broken.' The specific connotation of
this complex term is set by me ; I, the organism, select a definite
project complex, embodying definite implicative, discriminatory
relations (such as: 'a horse which my boys can ride,' 'a horse
that is not afraid of trolley cars,' etc.). Now, what of the extra-
cognitive complexes, of which this may be the projection? They
are indefinitely numerous, and there is in the projection itself abso-

HALLUCINATIONS 461

lutely no indication limiting their number. They are, with refer-
ence to my 'idea,' projective indiscernibles.

5. The alleged mental character of hallucinatory objects. To phi-
losopher and ordinary man alike, these entities have always been
stumbling blocks in the way to radical realism. And realists have
done little to remove them, prior to the recent courageous analyses
of Alexander and Nunn. The position these investigators are now
defending is, I believe, fundamentally correct; and when people
thoroughly understand it, they will wonder why everybody hadn't
discovered it centuries ago. They are far from understanding it,
though. Even such an acute reader as Lovejoy finds in it only a
subtle flouting of the unchallengeable facts. For this two circum-
stances are responsible : the first is that the naive realist's difficulty
here is traceable to his presuppositions, and of these he is not
clearly aware; the second is that Alexander and Nunn have thus
far given us no explanation, no description of the structure of a
universe in which hallucinatory objects can arise, and so their
hypothesis seems to be a mere ' blanket.' To made this appear-
ance all the worse, they treat only the stuff of hallucinatory ob-
jects as real, leaving the erroneous meanings more or less products
of a construing mind. This is all too perilously like old-fashioned
dualism, to please any realist. It must be displaced by an inter-
pretation which makes equivocal values and ' misconstructions ' of every
sort no less independent of cognition than true propositions are.
Now, the first step in this direction is the clearing up of the two
circumstances I have alluded to. This I shall endeavor to do by
scrutinizing the common-sense verdict about hallucinations and
its underlying presuppositions. Lovejoy has admirably summa-
rized that verdict as follows:

"Spatial objects may at least in some cases really exist
in some realm or medium other than that of real space. Such a
realm or medium, so far as I can see, is precisely what people ordi-
narily mean by 'consciousness' ; and the kind of object that has
its subsistence therein is what they ordinarily mean by an 'image'
or 'representation.' But the existence of an object in this me-

462 IMPLICATIONS OF BIOLOGY

dium evidently is not properly describable as the momentary en-
trance of a real and perduring spatial thing into a new relation
with other things ; for in the case of the hallucination the particu-
lar thing that is 'in consciousness' does not perdure and does not,
though perceived as spatial, exist in the same real space in which
other objects are still — by the new realist and by common sense
— supposed to exist." l

Lovejoy here defends a view which is inevitable so long as the
cognitive situation is not viewed as the projective field of a projec-
tion system. From this new standpoint, however, his opinion is
seen to be based on the two following implicit fallacies :

1. "The entity A enters a new relation r ; therefore every char-
acter of A enters this same relation identically." This is the
necessary presupposition of the statement that the hallucinatory
object cannot be the real object entering into new relations be-
cause the former does not possess all the characters of the latter.
For, clearly, if only a single feature of a thing could enter into a
given relation, and if the thing, in that relation, could manifest only
one of its properties, then there would be ground for Lovejoy's
inference.

Now, how about the presupposition ? Can it be defended ? I
do not see how, for it is an extreme variety of our old enemy, the
internal relation theory. Virtually it identifies the thing-in-rela-
tion with the thing-in-itself . It alleges that all the qualities and
relations of a thing are so tightly knit together 'organically' (in
the metaphysical, unbiological sense, of course) that all of them
must equally share in each new relation into which the thing en-
ters. If some of them do not, then the ' real ' thing is not in the
given relation ; there is an 'appearance,' a 'mental state' in its
stead. The instant you test such a hypothesis, however, its im-
possibility becomes visible. It means that a bay horse cannot
be physically related to the camera which photographs him be-
cause the animal, in relation to the camera, is only a few inches

1 Lovejoy, Reflections of a Temporalist on the New Realism, /. of Phil., Psychol.,
etc., 8, 596.

HALLUCINATIONS 463

long, shows only one eye, is only gray, etc. Or if an object enters
a magnetic field, all its qualities must be in the magnetic relation,
hence, its color, flavor, shape, weight, texture, market value, and
a thousand other features are all magnetized. If the physicist can-
not find them in this relation, by the aid of delicate registering
instruments, the thing itself is not there.1

2. "A hallucinatory complex does not exist at that place in
space where it appeal's to exist ; therefore it does not exist in real
space anywhere." This fallacy is based upon the unnoticed pre-
supposition that consciousness is the mere knowing of physical
things, as these are 'in themselves.' Of course, if one clings to
this naive doctrine, the fact of error and hallucination leads irre-
sistibly to a dualism. But once regard both the physical things
and their cognized phases as relational complexes, and the whole
perplexity vanishes, simply by your recognizing that the character
of a space complex, such as a man seen in a hallucination, is physi-
cally determined by its relations to many other entities. In this
respect, space characters do not differ at all from color characters.

A hillside which is green near at hand, under a certain light, is
bluish green a little farther off, bluish gray at a greater distance, and
blue from a still remoter vantage point. That these are its real
physical colors the spectroscope proves. From this fact, we readily
infer to-day that color is not a mysterious essence inside of the hill,
but a character which is a function of many things, ether, air, angles,
distances, etc. Now is it not obvious that a character which is
constituted by such a complex cannot be located in any one part of
that complex exclusively? You might as well try to put the cir-
cularity of a circle into one of the points of the circumference. Of
course, what you can do is to discern in the relation of such a point
to other points a peculiarity which, given a sufficient linear
continuum, will constitute a circumference. And this is the very
thing you must do with the color of the distant hill. Color is

1 This fallacy underlies Bergson's theory of cognition, in which he maintains that
knowledge never gives us ' the real thing ' inasmuch as it gives us only a few of
its selected characters.

464 IMPLICATIONS OF BIOLOGY

not in a point, as though it were an 'essence' whose 'appearances'
emanated therefrom. Color is a character of a considerable ex-
tent of ether which is disturbed in a certain manner. Like all
other physical characters, it varies with the number, arrangement,
and character of ether units and other things; and, in the sense
which Perry has defined the term 'dependence,' the color depends
upon the total complex either as an effect upon its exclusive cause
or as the whole upon its own parts. Some reasonably naive people
can grasp this and assent to it. But the old essence-substance
notion troubles them when they try to think in the same manner
of space characters. How about the size and shape of the hill ?
Is it not many thousand times as big as a man ? And shaped like
a sugar loaf? And at just one place in the universe, at a given
instant ? I grant that one must be quite sophisticated to answer
these queries in a thoroughly relationalistic manner. We have all
become so accustomed to construe space in terms of measure that
every other interpretation of it sounds absurd. Our practical
dealings with space are all metrical, of course. The important
questions of daily life are : How far ? How near ? How much to
the right? How many degrees to the left? And they are an-
swered largely by our muscles, which carry our limbs across the
appropriate stretches. Now, there is no denying that one posi-
tion and only one lies at a given distance and direction from another
position. Hence, if the relation between the two positions were
exclusively that of distance and direction, in the proper quantita-
tive sense, we should have to agree with the popular view that an
object which 'seems' to occupy a position which is not its 'real'
one, in terms of measure, is 'not really' the latter object at all, but
only a phantasm of the brain. That this opinion has prevailed
among philosophers no less than among laymen is due to a mere
motor instinct, . coupled with ignorance of modern geometry and
the profounder types of space relations.

As soon as we construe elements in the cognitive field as a pro-
ject-complex, we detect the error in the inference that, because
a space object that is cognized has not the same metrical values as

REAL INDISCERNIBLES 465

the object which it purports to be, therefore it is (a) not in real
space and (6) not the real object. The fallacy rests upon the pre-
supposition that two entities which are indiscernibles in one rela-
tion and discernibles in another cannot both fall within the con-
tinuum to which one of them indisputably belongs. That this
is false may be proved, not only by formal logic, but still more
convincingly by an appeal to a concrete case in which it does not
hold. By formal logic : A is a member of the class M, and in this
relation (alone) it is indistinguishable from B, which is also a mem-
ber of M ; now, if A is distinguishable f roin B in some other rela-
tion than that of membership in M, we cannot infer that A is not
(a) a member of M nor (6) a part of B, in case B is a complex.
Such an inference would be the ordinary fallacy of accident. This
can be shown concretely in projective geometry. Here all pairs
of points on a given line are indistinguishable in themselves; that is,
so long as other points and lines are not determinately given and
certain operations performed. In other words, their distances
are indiscernible, hence the 'absolute locus' of each point. But
from this fact, does the geometer infer that the point A of the line
M cannot be 'really' in M, because A is projectively indistinguish-
able from B, of the line M ? Or that A is a member of some other
class N? Not at all. He admits the obvious, namely, that en-
tities in certain relations are real indiscernibles, and that their iden-
tity in such complexes as are determined only by those relations is
just as genuine as their differences in complexes otherwise determined.
Were the philosopher to be no less scrupulous, he must say that
optical and other sensory illusions, hallucinations, and delusions
are not entities in some other continuum than that in which they
exist ' immediately,' but rather entities so related to others in that
continuum that they are, in that particular relation, identical in
quality, efficiency, and every other respect. Thus, there is a
class of complex entities in space and time so related hi one respect
(which I should like to call a 'direction,' but for the narrow mean-
ing popularly read into the term) that the complex of members A,
B, C, . . . is identical with the complex of members M, N, 0, . . .

2H

466 IMPLICATIONS OF BIOLOGY

The former complex is, when described in its other, more inclusive
relations, a stone wall in a given place ; the second complex is a
man who, at the given moment, is eating his breakfast a hundred
miles away from the stone wall. I, having a hallucination, see the
man where the stone wall 'really' is; but this fact does not prove
that either the man or the wall is not in real space : it is only an
illustration of the fact that two spatial things may be protectively
related in a cognitive field not less than in a field of lower order,
so that they are there identical (indiscernible).

6. Conclusion. — Here our formal analysis of the biological situa-
tion must halt, for lack of space in which to describe the other fac-
tors, such as the cognizing reagent, the motor response to relations
in the cognitive field, and the mechanism by which this response
is effected. Short as our journey has been, though, we have come
upon not a few important facts. The realistic implications of
ordinary biological description clear up, in large measure, the
problem of error, which is the central issue in every theory of con-
sciousness. Practically every argument for a separate realm of
consciousness or spirit rests its case ultimately upon the appar-
ently certain fact that many things ' in ' the mind do not exist in
the external world, although they do seem to. This indiscernibil-
ity of seeming from being, in the field of immediate experience,
has almost invariably been construed as proving that, in immedi-
ate experience, it is not the 'things in themselves' we are dealing
with, but entities of some other and unique order. Now, we have
seen that (a) the indiscernibility of seeming from being can be
naturally attributed not to the terms but to the relation in which
they stand; and (6) this relation is of a genus familiar to the
modern geometer ; and (c) exists in the space order ; and (d) is not
confined to space but is common to 'forms of externality' or dimen-
sional structures.

From these facts the biologist is free to draw an inference which
has extremely high probability, namely, that organisms have
developed elaborate projection systems in the central nervous
systems just because the physical world-order is so full of charac-

ERROR 467

ters which, with reference to the position occupied by a given or-
ganism at any given instant, are indiscernibles and yet, in other
relations, are different from one another. At each moment a
factor of the environment must be given in several projections, or
(what amounts to the same thing) a sufficient number of distinct
elements must be given, to make the character as a whole dis-
cernible. This conjecture, which we cannot here set forth in de-
tail, leaves us with a view of error exactly opposite to that tradi-
tionally championed. Error is not a product of the nervous system,
but the nervous system is a contrivance to deal with a physical state
of affairs of which error is only a very intricate instance.

APPENDIX

APPENDIX

THE PROGRAM AND FIRST PLATFORM OF SIX REALISTS1

PHILOSOPHY is famous for its disagreements, which have contrib-
uted not a little towards bringing it into disrepute as being unscientific,
subjective, or temperamental. These disagreements are due in part,
no doubt, to the subject matter of philosophy, but chiefly to the lack of
precision and uniformity in the use of words and to the lack of deliberate
cooperation in research. In having these failings philosophy still differs
widely from such sciences as physics and chemistry. They tend to make
it seem mere opinion ; for through the appearance of many figurative or
loose expressions in the writings of isolated theorists, the impression is
given that philosophical problems and their solutions are essentially per-
sonal. This impression is strengthened by the fact that philosophy con-
cerns itself with emotions, temperaments, and taste. A conspicuous
result of this lack of cooperation, common terminology, and a working
agreement as to fundamental presuppositions is that genuine philo-
sophical problems have been obscured, and real philosophical progress
has been seriously hindered.

It is therefore with the hope that by cooperation genuine problems
will be revealed, philosophical thought will be clarified, and a way opened
for real progress, that the undersigned have come together, deliberated,
and endeavored to reach an agreement. Such cooperation has three
fairly distinct, though not necessarily successive stages: first, it seeks
a statement of fundamental principles and doctrines; secondly, it aims
at a program of constructive work following a method founded on these
principles and doctrines ; finally, it endeavors to obtain a system of axioms,
methods, hypotheses, and facts, which have been so arrived at and formu-
lated that at least those investigators who have cooperated can accept
them as a whole.

After several conferences the undersigned have found that they hold
certain doctrines in common. Some of these doctrines, which constitute
a realistic platform, they herewith publish in the hope of carrying out

1 Reprinted from the /. of Phil., Psychol, etc., 1910, 7, 393.
471

472 APPENDIX

further the program stated above. Each list has a different author, but
has been discussed at length, revised, and agreed to by the other conferees.
The six lists, therefore, though differently formulated, are held to repre-
sent the same doctrines.

By conferring on other topics, by interchange of ideas, and by sys-
tematic criticism of one another's phraseology, methods, and hypotheses,
we hope to develop a common technique, a common terminology, and so
finally a common doctrine which will enjoy some measure of that author-
ity which the natural sciences possess. We shall have accomplished one
of our purposes if our publications tempt other philosophers to form small
cooperative groups with similar aims.

EDWIN B. HOLT, Harvard University.
WALTER T. MARVIN, Rutgers College.
W. P. MONTAGUE, Columbia University.
RALPH BARTON PERRY, Harvard University.
WALTER B. PITKIN, Columbia University.
E. G. SPAULDING, Princeton University.

1. The entities (objects, facts, etc.) under study in logic, mathematics,
and the physical sciences are not mental in any usual or proper meaning
of the word 'mental.'

2. The being and nature of these entities are in no sense conditioned by
their being known.

3. The degree of unity, consistency, or connection subsisting among
entities is a matter to be empirically ascertained.

4. In the present stage of our knowledge there is a presumption in favor
of pluralism.

5. An entity subsisting in certain relations to other entities enters into
new relations without necessarily negating or altering its already sub-
sisting relations.

6. No self-consistent or satisfactory logic (or system of logic) so far
invented countenances the 'organic' theory of knowledge or the 'inter-
nal' view of relations.

7. Those who assert this (anti-realistic) view, use in their exposition

a logic which is inconsistent with their doctrine.

EDWIN B. HOLT.

PROGRAM AND FIRST PLATFORM 473

II

1. Epistemology is not logically fundamental.1

2. There are many existential, as well as non-existential, propositions
which are logically prior to epistemology.2

3. There are certain principles of logic which are logically prior to all
scientific and metaphysical systems.

One of these is that which is usually called the external view of relations.

4. This view may be stated thus : In the proposition, "the term a is in
the relation R to the term b," aR in no degree constitutes b, nor does Rb
constitute a, nor does R constitute either a or 6.

5. It is possible to add new propositions to some bodies of information
without thereby requiring any modification of those bodies of information.

6. There are no propositions which are (accurately speaking) partly
true and partly false, for all such instances can be logically analyzed into
at least two propositions one of which is true and the other false. Thus
as knowledge advances only two modifications of any proposition of the
older knowledge are logically possible ; it can be rejected as false or it can
be analyzed into at least two propositions one of which is rejected.

1 Some of the principles of logic are logically prior to any proposition that is de-
duced from other propositions. The theories of the nature of knowledge and of the
relation of knowledge to its object are for this reason logically subsequent to the prin-
ciples of logic. In short, logic is logically prior to any epistemological theory.
Again, as theories of reality are deduced and are made to conform to the laws of
logic they too are logically subsequent to logic ; and in so far as logic is logically
present in them it is itself a theory or part of a theory of reality.

2 The terms 'knowledge,' 'consciousness,' and 'experience' found in common sense
and in psychology are not logically fundamental, but are logically subsequent to
parts at least of a theory of reality that asserts the existence of terms and relations
which are not consciousness or experience. E.g. the psychical is distinguished
from the physical and the physiological.

Now idealism has not shown that the terms 'knowledge,' 'consciousness,' and
'experience' of its epistemology or of its theory of reality are logically fundamental
or indefinable, nor has it succeeded in defining them without logically prior terms
that are elsewhere explicitly excluded from its theory of reality. In short, idealistic
epistemologists have borrowed the terms ' knowledge, ' ' consciousness, ' and ' ex-
perience from psychology, but have ignored or denied the propositions in psychology
that are logically prior. In other words, epistemology has not thus far made itself
logically independent of psychology nor has it freed itself logically from the com-
mon-sense dualism of psychology. On the contrary, epistemology from Locke
until to-day has been and has remained, in part at least, a branch of psychology.

474 APPENDIX

As corollaries of the foregoing :

7. The nature of reality cannot be inferred merely from the nature of
knowledge.

8. The entities under study in logic, mathematics, physics, and many
other sciences are not mental in any proper or usual meaning of the word
' mental.'

9. The proposition, "This or that object is known," does not imply that
such object is conditioned by the knowing. In other words, it does not
force us to infer that such object is spiritual, that it exists only as the ex-
periential content of some mind, or that it may not be ultimately real

just as known.

WALTER T. MARVIN.

ni

I. The Meaning of Realism.

1. Realism holds that things known may continue to exist unaltered
when they are not known, or that things may pass in and out of the cogni-
tive relation without prejudice to their reality, or that the existence of a
thing is not correlated with or dependent upon the fact that anybody ex-
periences it, perceives it, conceives it, or is in any way aware of it.

2. Realism is opposed to subjectivism or epistemological idealism which
denies that things can exist apart from an experience of them, or inde-
pendently of the cognitive relation.

3. The point at issue between realism and idealism should not be con-
fused with the points at issue between materialism and spiritualism,
automatism and interactionism, empiricism and rationalism, or pluralism
and absolutism.

II. The Opposition to Realism. Among the various classic refutations
of realism the following fallacious assumptions and inferences are promi-
nent :

1. The Psychological Argument : The mind can have for its direct ob-
ject only its own ideas or states, and external objects, if they exist at all,
can only be known indirectly by a process of inference, of questionable
validity and doubtful utility. This principle is fallacious because a know-
ing process is never its own object, but is rather the means by which some
other object is known. The object thus known or referred to may be
another mental state, a physical thing, or a merely logical entity.

2. The Intuitional Argument : This argument stands out most promi-

PROGRAM AND FIRST PLATFORM 475

nently in the philosophy of Berkeley. It has two forms. The first con-
sists of a confused identification of a truism and an absurdity. The tru-
ism : We can only know that objects exist, when they are known. The
absurdity : We know that objects can only exist when they are known. The
second form of the arguments derives its force from a play upon the word
' idea,' as follows : Every 'idea' (meaning a mental process or state) is incap-
able of existing apart from a mind; every known entity is an 'idea' (mean-
ing an object of thought) ; therefore, every known entity is incapable of ex-
isting apart from a mind. It is to the failure to perceive these fallacies
that ideah'sm owes its supposedly axiomatic character.

3. The Physiological Argument: Because the sensations we receive
determine what objects we shall know, therefore the objects known are
constructs or products of our perceptual experience. The fallacy here
consists in arguing from the true premise that sensations are the ratio
cognoscendi of the external world, to the false conclusion that they are
therefore its ratio fiendi or essendi.

III. The Implications of Realism.

1. Cognition is a peculiar type of relation which may subsist between a
living being and any entity.

2. Cognition belongs to the same world as that of its objects. It has
its place in the order of nature. There is nothing transcendental or super-
natural about it.

3. The extent to which consciousness pervades nature, and the condi-
tions under which it may arise and persist, are questions which can be
solved, if at all, only by the methods of empiricism and naturalism.

W. P. MONTAGUE.

IV

1. The object or content of consciousness is any entity in so far as it is
responded to by another entity in a specific manner exhibited by the reflex
nervous system. Thus physical nature, for example, is, under certain
circumstances, directly present in consciousness.

In its historical application, this means that Cartesian dualism and the
representative theory are false ; and that attempts to overcome these by
reducing mind and nature to one another or to some third substance, are
gratuitous.

2. The specific response which determines an entity to be content of

476 APPENDIX

consciousness does not directly modify such entities otherwise than to
endow them with this content status. In other words, consciousness
selects from a field of entities which it does not create.

In its historical application, this implies the falsity of Berkeleian and
post-Berkeleian idealism in so far as this asserts that consciousness is a
general ratio essendi.

3. The response which determines an entity to be content may itself
be responded to and made content in like manner. In other words, the
difference between subject and object of consciousness is not a difference
of quality or substance, but a difference of office or place in a configura-
tion.

In its historical application, this implies the falsity not only of the Car-
tesian dualism, but of all idealistic dualisms that, because they regard
subject and object as non-interchangeable, conclude that the subject is
either unknowable, or knowable only in some unique way such as intui-
tively or reflexively.

4. The same entity possesses both immanence, by virtue of its mem-
bership in one class, and also transcendence, by virtue of the fact that it
may belong also to indefinitely many other classes. In other words, im-
manence and transcendence are compatible and not contradictory predi-
cates.

In its historical application, this implies the falsity of the subjectivistic
argument from the ego-centric predicament, i.e. the argument that be-
cause entities are content of consciousness they can not also transcend
consciousness ; it also implies that, so far as based on such subjectivistic
premises, the idealistic theory of a transcendent subjectivity is gratuitous.

5. An entity possesses some relations independently of one another;
and the ignorance or discovery of further relations does not invalidate a
limited knowledge of relations.

In its historical applications, this implies the falsity of the contention
of absolute idealism that it is necessary to know all of an entity's relations
in order to know any of its relations, or that only the whole truth is wholly
true.

6. The logical categories of unity, such as homogeneity, consistency,
coherence, interrelation, etc., do not in any case imply a determinate de-
gree of unity. Hence the degree of unity which the world possesses can
not be determined logically, but only by assembling the results of the spe-
cial branches of knowledge. On the basis of such evidence, there is a

PROGRAM AND FIRST PLATFORM 477

present presumption in favor of the hypothesis that the world as a whole
is less unified than are certain of its parts.

In its historical application, this implies that the great speculative mon-
isms, such as those of Plato, Spinoza, and certain modern idealists, are
both dogmatic and contrary to the evidence.

RALPH BARTON PERRY.

The realist holds that things known are not products of the knowing
relation nor essentially dependent for their existence or behavior upon
that relation. This doctrine has three claims upon your acceptance:
first, it is the natural, instinctive belief of all men, and for this, if for no
other reason, puts the burden of proof upon those who would discredit
it ; secondly, all refutations of it known to the present writer presuppose
or even actually employ some of its exclusive implications ; and, thirdly,
it is logically demanded by all the observations and hypotheses of the
natural sciences, including psychology.

Involved more or less intimately in a realistic view are the following :

1. One identical term may stand in many relations.

2. A term may change some of its relations to some other terms with-
out thereby changing ah1 its other relations to those same or to other terms.

3. What relations are changed by a given change of relation cannot
always be deduced merely from the nature of either the terms involved
or their relation.

4. The hypothesis that 'there can be no object without a subject' is
pure tautology. It is confessedly a description of the cognitive situation
only; and it says, in effect, that everything experienced is experienced.
It becomes significant only by virtue of the wholly unwarranted assump-
tion that doctrines 1, 2, and 3, above given, are false. This assumption,
however, is fatal to the idealist's supposed discovery, inasmuch as it means
that there can be no true propositions. In conceding this, the idealist
refutes himself.

5. In no body of knowledge, not even in evidences about the nature of
the knowledge relation, can we discover that possible knowledge is limited
or what its limits may be.

6. Entities are transcendent to the so-called 'knowing mind' or 'con-
sciousness ' only as a term is to the relations in which it may stand, viz.

478 APPENDIX

in two radically different manners : first, as the term is not identical with
the particular relation in which it stands, so too a thing in the knowledge
relation is not the relation itself ; secondly, as the term may enter into or
go out of a particular relation, without thereby being changed essentially
or destroyed, so too can an object of knowledge exist prior to and after
its entrance into or removal from the knowledge relation. Transcend-
ence thus means, in the first place, distinctness and, in the second place,
functional independence.

7. There may be axiomatic truths or intuitive truths. But the fact
that a truth belongs to either of these classes does not make it fundamental
or important for a theory of knowledge, much less for a theory of reality.
Like all other truths, it too must be interpreted in the light of other rele-
vant truths.

8. Though terms are not modified by being brought into new contexts,
this does not imply that an existent cannot be changed by another ex-
istent.

WALTER B. PITKIN.

VI

1. Realism, while admitting the tautology that every entity which is
known is in relation to knowing or experience or consciousness, holds that
this knowing, etc., is eliminable, so that the entity is known as it would
be if the knowing were not taking place. Briefly, the entity is, in its being,
behavior, and character, independent of the knowing. This position
agrees with common sense and with science in holding (1) that not all
entities are mental, conscious, or spiritual, and (2) that entities are know-
able without being known.

2. The fact that terms are in the cognitive relation does not imply that
the terms are mutually dependent on, or capable of modifying, either each
other or the relation, any more than this dependence, etc., is implied for
any two terms in any other relation. The proposition that there is this
dependence, etc., constitutes the 'internal view' of relations.1 Most of

1 To hold the 'internal view' means, in my opinion, to hold that, in order that
a relation may relate, the relation must either (1) penetrate its terms, or (2) be
mediated by an underlying (transcendent) reality. From the penetration there
is deduced (a) modification, or (6) similarity, or (c) the generation of a contradic-
tion. Cf. my paper, The Logical Structure of Self-refuting Systems, Phti. Re-
view, 19, 277-282.

PROGRAM AND FIRST PLATFORM 479

those systems which are opposed to realism can be shown to presuppose
this 'internal view,' but this view can be shown to be self-contradictory
and to presuppose the 'external view.'

3. That position which is based in part on the acceptance and the con-
sistent use and development of the implications of these logical doctrines
which are presupposed as a condition for any position being stated, argued,
and held to be true, has, thereby, a strong presumption created in favor
of its truth.1

4. There is at least one logical doctrine and one principle which are
ultimately presupposed by any system which is held to be true. That
doctrine is the 'external view' of relations, and the principle is that
truth is independent of proof, although proof is not independent of truth.
The first of these means, briefly :

5. (1) That both a term and a relation are (unchangeable) elements or
entities ; (2) that a term may stand in one or in many relations to one or
many other terms ; and (3) that any of these terms and that some of
these relations could be absent or that other terms and relations could be
present without there being any resulting modification, etc., of the re-
maining or already present terms or relations.

6. By this 'external view' it is made logically possible that the know-
ing process and its object should be qualitatively dissimilar. (Cf. 1.)

7. The principle (cf . 4) means, that, while on the one hand no proposi-
tion is so certain that it can be regarded as exempt from examination,
criticism, and the demand for proof, on the other hand, any proposition,
if free from self-contradiction, may be true (in some system). In this
sense every proposition is tentative, even those of this platform.

Corollary. — It is impossible to get a criterion, definition, theory, or con-
tent for the concept 'absolute' by which it can be absolutely known or
proved that any criterion, definition, theory, or content is absolutely true,
i.e. is more than tentative. The most that can be claimed for such a
criterion, etc., is that it may be absolutely true, although not proved to be.

1 Such a system / hold to be realism, its chief feature being the interpretation of
the cognitive relation in accordance with the 'external view.' This 'external
view ' can be held to be true quite consistently with itself, and is in this sense, I hold,
self-consistent, as is also, in my opinion, realism. Accordingly I hold further that
realism is not a merely dogmatic system, and that, as self-consistent, it refutes and
does not merely contradict certain opposed systems which, as based on the ' internal
view,' are self -refuting.

480 ILLUSORY EXPERIENCE

8. Any entity may be known as it really is in some respects without its
being known in all respects and without the other entities to which it is
related being known, so that knowledge can increase by accretion.

9. Knowing, consciousness, etc., are facts to be investigated only in the
same way as are other facts, and are not necessarily more important than
are other facts.

10. The position stated in this platform, which is a position concerning
knowing as well as other things, can apply to itself, as a special instance
of knowledge, all its own propositions about knowledge.1

EDWABD GLEASON SPAULDING.

NOTE ON PROFESSOR HOLT S ESSAY

WITH a view to a clearer understanding by the reader of the theories of
error and consciousness set forth by Holt and myself, I here summarize
what seem to me to be the main points at issue between us :

I. I disagree with Holt's theory that specific qualities are mere temporal
condensations of primitive and qualitatively simple pulses, on the ground
that such temporal condensation constitutes the attribute of intensity.
When hi a series of stimuli there occur positive intervals of empty tune
between their effects on the sensory center, we experience the series as a
succession of discretes. If the series of stimuli increases in rapidity to
the point when the intervals between their effects on the sensory center
fall to zero, we experience the series as a single continuously enduring
quality. When the succession of stimuli becomes so rapid that their effects
upon the sensory center actually overlap or interpenetrate, we get not a
new quality but an increase in the original intensity of the old quality —
an increase which is proportioned to the degree of the overlapping or tem-
poral condensation. In short, Holt's concept of temporal condensation
provides a good analytic definition of the category of intensity, and for
that very reason it cannot be used to express qualitative difference.

II. I disagree with Holt's doctrine that contradictions are objective
and related after the manner of opposing forces, and hence with his con-
clusion that these objective contradictions constitute the content of an

1 1 hold that for this reason the position here stated is self-critical, and that it is
this which distinguishes it from a large class of historical systems, notably phenome-
nalism, subjective and objective idealism, and absolutism.

SPECIFIC ENERGIES 481

erroneous experience and cause its occurrence. The unreal object or con-
tent of an error subsists extra-mentally but it does not contribute in any
causal manner to its being apprehended. It is the nature of the unreal,
or merely subsistent, to be sterile of consequences. It can be known but
it cannot cause itself to be known, and apart from being known it has no
efficiency. The non-existent bent stick cannot cause us to perceive it,
but an existent straight stick, partly immersed in water, can produce by
reason of the different refracting powers of water and air an effect upon the
eye and brain of the same kind as would have been produced by a stick
that was really bent, with the result that a non-existent bent stick becomes
the object of an (erroneous) apprehension.

III. The most serious difference between Holt's conclusions and my
own concerns the nature of consciousness. We both agree that extra-
organic objects produce in the brain effects which more or less resemble
their causes and which are the means by which those causes are perceived.
But we differ as to the manner in which these brain effects are instrumental
to perception. In the first part of Holt's paper, he appears to me to ad-
duce evidence that would support the view that the objects of which we
are conscious are within the brain. He speaks (1) of the effects produced
in the nervous system as being true parts of the extra-organic objects
which cause them, and of the object being present in its effects, as a man's
expression is truly present in his portrait; (2) of the temporal condensa-
tions of neural vibrations as the basis for the experience of secondary
qualities; (3) of the non-psychical character of illusory objects, as im-
plied by their similarity to the distorted images on a photographic plate.

It seems to me that these arguments, if valid, would obviously imply
that the objects of which we are conscious constitute the system of effects
projected upon the brain; and that, when this conclusion is disavowed the
foregoing evidence loses all its relevancy. As I understand it, it is this
conception of consciousness as consisting of the intra-organic projections
of extra-organic objects that is defended by Pitkin in his article. Now
despite the plausibility of the projection theory it is, I believe, open to
certain insuperable objections which however need not be mentioned here,
for it seems that after all Holt rejects it as explicitly as 1 do, and in the
last part of his article sets forth what appears to me the wholly different
conception of consciousness as a 'specific response' to extra-organic
objects which by virtue of being thus responded to become the field of

objects in consciousness.
2i

482 ILLUSORY EXPERIENCE

The whole question of the nature of consciousness thus seems to hinge
on the meaning of the 'specific response.' But I cannot find any ade-
quate statement of what this response is. In spite of its alleged 'specific'
character, consciousness, it seems to me, might as well be identified with
X. If by the response Holt means anything physical, such as an organic
movement or efferent nerve current, my objections would be those set
forth in my article against the theory that consciousness is behavior.
If, on the other hand, the specific response which constitutes consciousness
is not behavior or movement, the only alternative seems to me to be
that of identifying it with the relation of self-transcending implication,
which the brain-states sustain to their extra-organic causes. But this
is the view which I have defended and which Holt rejects.

W. P. MONTAGUE.

NOTE ON PROFESSOR MONTAGUE'S ESSAY

As the reader of our two essays will readily gfee, my view of consciousness
differs widely from that of Montague. A fundamental feature of this
difference is to be found in the concept of self-transcendent reference,
on which Montague's theory hinges; for, according to him, "it is the
self -transcending implications of these brain-states that constitute our
consciousness of the spatio-temporal world in which we live." And the
germ of such self-transcendence seems to be found in the causal rela-
tion: "Each event-element has over and above its own qualities and its
own position in space and time, something which implies or refers to other
events. It is both an agent and a patient of what is not itself." Now
whether or not a case of self-transcendent reference is anywhere to be
found, I do not think that one is found in the relation between brain-
states and their correlated contents of consciousness. The brain-state,
as I believe, does not refer by self-transcendence to the object that is at
the moment content of consciousness, but it is a specific response to that
object. Nor, again, do I find such self-transcendence in the case of
causality, for here I discover nothing but the constant function of an
independent variable (generally time).

Again, Montague and I differ in our understanding of error. I do not
look on error as primarily or even generally attributable to anything that
could be called the 'distorting' influence of the physiological mechanism
of perception and thought. It seems to me that the extra-mental world is

SPECIFIC ENERGIES 483

teeming with contradictions and unrealities, and that these can come to
consciousness by virtue of a psychical process, which presents no elements
of 'distortion.'

The application of Montague's view and of mine to the empirical facts
of psychology, and of modern logic, must decide as to which is the better

descriptive and explanatory principle.

E. B. HOLT.

NOTE ON THE ESSAYS OF PROFESSORS HOLT AND MONTAGUE

THERE is much in these two essays which I should like to discuss,
if space permitted. Holt's theory of time-density and of cognitive cross-
sections and Montague's rejection of the projective hypothesis interest
me most acutely. I agree fundamentally with Holt in his two doctrines,
though I am not at all sure that time-density yields qualitative characters.
It may well be, as Montague suggests, that it yields only the intensity
series. This is a matter calling for thorough investigation, and I hope
we may soon attack it. But Holt's underlying contention is still sound.
The alleged simple qualities and their unique specificities are ' form qualities '
and their simple constituents are minute physical events. These events
are packed together, now in space, now in tune, and again, perhaps, in
both continua. If time-density yields intensity, space-density probably
yields quality ; but touching this, I confess to much ignorance.

As for Montague's rejection of the projective hypothesis, I believe
it is due to his supposing that projection is the simple casting of simulacra
upon a screen. This is a very natural misunderstanding, for all cases
of spatial projection do involve the generating of new figures definitely
related to the projected figures. But these new figures are not the pro-
jective relation. They are only the terms of the relation. And it is just
this, their relation to other terms, which constitutes the peculiarity called
error, or, more precisely, multiple value or indiscernibility.

Montague, I believe, opposes the projective hypothesis chiefly on
the ground that it leaves the sentient organism in possession of nothing
more than its own brain states, to wit, the images projected upon some
parts of the nervous system. Now, there are such images, at least in
the more finely organized sensory systems ; but they are incidental products
of the projective relation, and not the significant nature of it. The efficient
projective relations are not, as it were, inside the individual sensory
elements ; they obtain between the latter, or, in other cases, between the
organism (its position or condition) and the external entities reacted to.
In short, they are not merely odd chemical properties of the cortex ; they

484

APPENDIX 485

are not physical properties at all. Hence there is no danger of reducing
the mind to knowledge only of its brain states, inasmuch as these are
merely terms in the projective relation, and not the relation itself.

Montague and Holt and I agree that, whatever consciousness is, it
is somehow connected with the activity of getting 'beyond space and
time'; that is, of adjusting variously to events beyond the organism's
own skin and to conditions more than material. But I differ from Mon-
tague in locating the device by which this adjustment is effected. I
regard the implicative relation as transverse to the physical field of its
terms. The latter are in space and time; and their physical relations
are all spatio-temporal. But the implicative relation is not. It cuts
across the physical field, but it does not cut in any spatio-temporal dimen-
sion. (This is also Holt's opinion, if I mistake not.) Against this
view Montague looks upon the implicative relation as being longitudinal
to the physical field, actually in and through it, running from cause to
effect, and hence having a genuine historical flow. It is imbedded in the
stuff of things as electric charges are, as velocities, stresses and strains
and pulls are. It is, indeed, one phase of causation.

But for me this hypothesis engenders difficulties, of which I mention
only three :

1. Implication cannot be a mere counter-phase of causation, inas-
much as we find the implicative relation between terms which are not
causally related in any series. For instance, one unreal may imply
another: the death of the Emperor of the United States implies the
death of the head of the ruling American house. I know that Montague
denies that this is a genuine implication, as he denies all implying between
unreals. But to discuss this would be to open the whole field of modern
logic ; so I leave the reader with the problem and I pass to a more obvious
instance of non-causal implication, namely that between timeless entities.
The triangle implies a constant sum of interior angles; but the angles
are neither the cause nor the effect of the triangle, they are only parts
of it ; and the geometrical proposition is one about the whole implying
something about its parts. Again, one event may imply another which
is simultaneous with it and not causally connected. Thus sunlight in
New York City implies darkness in Hong Kong; here the two events
are effects of the same complex causes, not effects or causes of each other.

2. If causes and effects reciprocally imply each other, then the proximate
cause implies its effect, and w'ce versa. Hence, the last brain state before

486 APPENDIX

the cognizing of an object ought to be at least one of the implicates given
in that instant of consciousness. But this is not the case.

3. If, on the other hand, this implicative relation is transitive, the
percipient knows immediately some term in the implicative series beyond
his own brain state. But why does he know the one which, let us say,
is located at a given point in space? Why does he perceive the sun,
instead of an ether wave, two inches or eighteen thousand miles from
his retina? Why does the implication stop anywhere? Surely the
cause-effect series is prodigiously long, and the implicative relation is
demonstrably transitive ; hence the brain state might well be expected
to imply all its impliers, and thus simple perception ought to yield us the
longed-for all-in-all knowledge about the flower in the crannied wall.
The answer to this difficulty cannot be that an effect implies its simplest
cause only ; for the proximate cause is the simplest always, yet it is never
implied.

I raise these queries because I believe that they lose their awkwardness
as soon as we give up the longitudinal hypothesis and accept the transverse.
If I mistake not, my differences with Montague reduce almost wholly to
this issue.

WALTER B. PITKIN.

INDEX

INDEX

ACCELERATION, 204 ff.

ACCIDENTS, definition of, 264.

ACTIVITY, mental, 439.

ADJUSTMENT, 389.

AGREEMENT, necessity of explicit, 28 ff .

ALEXANDER, S., 439.

ALGEDONIC QUALITIES, 408.

ANALYSIS, 151, 155, 329-330, 338-339,
341-342 ; arithmetical, 173 ff. ; atomic,
225; and Bergson, 160 ff. ; concep-
tual, 230 ff. ; enumerative, 162 ff. ;
experimental, 156 ; formal, 156 ; intro-
spective, 332-333, 334-337, 343, 345-
349 ; as a necessary method, 24 ff. ;
perceptual, 230 ff.; and pragmatism,
158 ff. ; of secondary qualities, 331-349 ;
types of, 155.

ART, 141.

ATOMISM, psychological, 337-340, 350-
352.

ATOMS, 225 ff.

ATTENTION, limited field of, 292-294.

ATTRIBUTE-THING, 109.

BEING, 357-359, 372-373.

BELIEF, double meaning of, 256; nature

of, 292-294.
BERGSON, H., 160 ff., 171, 186, 206, 218,

224, 231 ff., 247, 379.
BERKELEY, G., 6 ff.
BIOLOGICAL SITUATION, formal analysis

of, 380.

BRADLEY, F. H., 160 ff.
BRAIN, are perceived objects inside ?

276-277.
BRENTANO, F., 331-337.

CAIRO, E., 17.

CAUSALITY, no place for in world of

pure fact, 264 ; meaning of, 265 ;

substantist view of, 266-267 ; posi-

tivist view of, 266-267.
CAUSATION, 109.
CHANGE OP STATE, 264.
CLASS, 170.
COGNITION, as selective rather than

creative, 295 ; as self -transcending

reference, 282-285.

COMPLEMENTARINESS OF COLORS, 312,

327, 334.

COMPLEXES, 119 ff.
CONCEPTS, 233-236.
CONDUCTION, 396.
CONSCIOUSNESS, 443; definition of, 281;

and nervous system, 353-355 ; no place

for in world of pure fact, 264.
CONTENT, mental, 147.
CONTRADICTION, 355-356, 361-365, 367,

370.
CRITICISM, cf. EPISTEMOLOGY; defined,

49; a vicious circle, 72 ff.
CROSS-SECTION, 354, 368, 373-375.

DEFINITION, 22 ff.

DEPENDENCE, and relation, 106.

DESCARTES, R., 4 ff.

DEWEY, J., 115, 149, 379, 437.

DOGMATISM, defined, 49 ff.

DRIESCH, H., 245 ff. ; 379.

DUALISM, 4 ff . ; futility of, 275 ; theory

of Epistemological, 251.
DURATION, 214 ff. ; Bergson's doctrine

of, 212 224; as function of space,

time, and quality, 263.
DYNAMICS, 212 ff.

ELEMENT OF CONSCIOUSNESS, 320, 337-
340, 350-352.

ENERGY, as stored in the brain, 285.

ENVIRONMENT, 443.

EPISTEMOLOGICAL TRIANGLE, 286-289.

EPISTEMOLOGY, cf. CRITICISM, DOGMA-
TISM, defined as the fundamental
science, 45 ff. ; its failure to contrib-
ute to our knowledge of reality,
83 ff. ; its logical position relatively
to the other sciences, 60 ff. ; its pre-
suppositions, 67 ff . ; not logically funda-
mental, 51 ff. ; presupposes a theory
of reality, 74 ff. ; its value to the met-
aphysician, 94 ff.

ERROR, 357 ff., 367 ff. ; 467, definition
of, 252 ; degrees of, 297-299 ; genesis
of, 289 seq. ; two kinds of, 298-292.

EUCKEN, R., 19.

EVENT, as function of space, time, and

489

490

INDEX

quality, 263 ; as ultimate element, 253,

254.

EXISTENCE, definition of, 255.
EXPERIENCE, equivocal use of, 250 ff. ;

meaning of, 150.

EXPERIMENTALISM, fallacies of, 259-260.
EXTENDED THING, 264. **
EXTERNAL RELATIONS, theory of, 167.

FALLACIES, definition by initial predica-
tion, 15 ff . ; egocentric predicament,
11 ff. ; exclusive particularity, 14 ff. ;
illicit importance, 19 ff. ; indiscernibles,
427; morphological, 434 : pseudo-sim-
plicity, 12 ff. ; psychological meton-
ymy, 256 ; psychophysical metonymy,
294-297 ; relation of to logic, 262 ;
speculative dogma, 16 ff . ; verbal sug-
gestion, 18 ff.

FALSE, definition of, 252 ; identity of
with unreal, 252.

FLATFISH, 397.

GEOTHOPISM, 390.

GESTALTQUALITAETEN, 340-341, 346-
349.

HALLUCINATIONS, 461.

HISTORY, 142.

HOBHOUSE, L. T., 131.

HUME, D., 7 ff.

HYLOPSTCHISM, definition of, 279 ; rela-
tion of to panpsychism and pan-
hylism, 280 ; relation of to positivism
and substantism, 279.

IDEALISM, theory of epistemological,251.

IDEAS, BERKELEY'S equivocal use of,
258.

IDENTITY, numerical, as function of space,
time, and quality, 263 ; qualitative, as
function of apace, time, and quality,
262.

ILLUSIONS, 303-311, 352-353.

IMPLICATES, perceived objects as, 282-
285.

IMPLICATION, 112, 441 ; relation of to
potentiality, 282-283.

IMPULSE, mediation of, 420.

INDEPENDENCE, cases of, 1 18 ff . ; and con-
sciousness, 126 ff.; of ego, 144,; and
knowledge, 126 ff. ; meanings of, 99 ff. ;
106 ff., 117; and non-relation, 113;
and perception, 149; and priority,
115.

INDISCERNIBLES, fallacy of, 427; pro-
jective, 458, 465.

JOACHIM, H., 104, 114, 128, 133.
JUDGMENT, double meaning of, 256-257.

KANT, I., 8 ff., 78 ff.

KNOWLEDGE, ambiguity of the word,

56 ff . ; the problem of its possibility,

61 ff., 70 ff.

LAWS OF THOUGHT, non-mental nature

of, 261.

LEAST PERCEPTIBLE DIFFERENCE, 417.
LOCKE, J., 4 ff., 101.
LOGIC, a fundamental science, 61 ff. ;

and consciousness, 129 ff . ; not a

science of the knowing process, 52 ff . ;

non-psychological nature of, 261-262.
LOGICAL RIGOR, necessity of, 25 ff.
LOVEJOY, A. O., 462.

McGlLVARY, E. B., 114.

MECHANICAL, how related to vital and
sensory, 284.

MEISLING, A. A., 312, 326-328, 349.

METAPHYSICS, its foundations, 92 ff., its
indebtedness to epistemology, 94 ff.

METONYMY, verbal fallacy of psycho-
physical, 256 ; material fallacy of
psychophysical, 294-297.

MONISM, futility of agnostic, 275-276.

MOORE, G. E., 146.

MOTION, analysis of, 193 ff. ; as function
of space, time, and quality, 264.

MUELLER, Johannes, 314, 330.

MUENSTERBERG, H., 316, 337-339.

MYSTICISM, 160.

NAIVE REALISM, theory of, 251.
NATURAL REALISM, theory of, 251.
NERVE-IMPULSE, periodicities of, 321-

325, 328, 349, 351.
NUMBERS, 173, 177 ff.

ORGANISMS, analysis of, 243 ff.

PANHYLISM, two types of, 268-269;

relation to behavior, 271 : refutation

of, 269-272.

PANPSYCHISM, refutation of, 272-275.
PART-WHOLE, 107.
PERCEPTION, 230 ff. and independence,

149 ; an ultimate knowledge, 64 ff.
PERSPECTIVE, Translations of, 401.
PHASE-UNITS, 254.
PHILOSOPHY, HISTORY OF, separation of

philosophy from, 30 ff.
POSITIVIST VIEW OF CAUSALITY, 266-

267.

POSTKANTIANISM, 8 ff.

INDEX

491

POTENTIALITY, directions of, 281-282;
levels of, 283-284; relation of to
consciousness, 281 ; relation of to
implication, 282-283.

PRAGMATISM, 158, 415, 437.

PRIORITY, logical, 204.

PROJECTION, 449 ff.

PROPOSITIONS, definiton of, 254 ; exist-
ential, 255 ; non-existential, 255 ;
universal, 255.

PROPOSITIONS vs. TERMS, 357-358, 262-
363 ; their subsistence, 57 ff. ; their
truth a logical ultimate, 59.

PSYCHOPHYSICAL METONYMY, material
fallacy of, 294-297.

QUALITIES, algedonic, 408; as ultimate
category, 263 ; primary and secondary,
308, 313, 355 ; secondary, 308 ff., 331
ff., 347, 352, 355.

RATIONALISM, 160.

REACTION, types of, 385 ; described,
384.

REAL, definition of, 255.

REALISM, analytical, 155 ; and biology,
39 ff., 377 ff. ; biological attack on,
378 ; constructive tendencies of, 31 ff.;
and epistemology, 41, 45 ff . ; mean-
ings of, 99, 103 ; naive, 2 ff., 251 ; New,
9 ff . ; polemic of, 11 ff. ; program of
the new, 21 ff.; and psychology, 37 ff.

REALITY, 304, 358-361, 366.

RECIPROCITY, 111.

REID, T., 100 ff.

RELATION, 165, and dependence, 106;
fallacy of internal, 299 ; internal theory
of, 165 ff.

RESISTANCE, 409.

REST, as function of space, time, and
quality, 263.

ROYCE, J., 103, 119, 125.

RUSSELL, B., 137, 166, 169, 170, 176, 182,
185, 187, 190, 202, 206, 210.

SCIENCE, metaphysics* indebtedness to

its progress, 84 ff.
SECONDARY QUALITIES, external reality

of, 299.

SELECTION, 393.

SIMILARITY, 337-339, as function of

space, time, and quality, 263.
SIMPLES, 118.
SOCIETY, 142.
SPACE, analysis of, 181 ff . ; continuity of,

184 ff . ; as ultimate category, 263.
SPECIFIC NERVE ENERGIES, 314-321,

328-330.

SPECIFIC RESPONSE, 354-355, 373.
SPENCER, H., 350.
STIMULUS, 383, 425.
SUBJECT, 144.

SUBJECTIVISM, 5 ff. ; theory of, 251.
SUBJECTIVITY and consciousness, 136 ff.
SUBSISTENT, definition of, 253 ; as in-
volving proposition, 253.
SUBSTANCE, 99 ff. ; as function of space,

time, and quality, 264 ; mental and

physical, 354, 368, 372.
SUBSTANCE-UNITS, 254.
SUBSTANTISM, types of, 267.
SUBSTANTIST, view of causality, 266-

267.

TERMS, 175; as changeless and duration-
less, 254.

THING-ATTRIBUTE, 109.

TIME, analysis, 190 ff. ; or ultimate
category, 263.

TRANSCENDENTALISM, its fallacious char-
acter, 75 ff.

TRANSFORMATION, 405.

TRUE, definition of, 252 ; identity with
real, 252.

TRUTH, definition of, 252 ; degrees of,
297-299 ; genesis of, 289 ff. ; two kinds
of, 289-292 ; Joachim on, 104.

VALUE, 140 ff., 148.
VELOCITY, 204 ff .
VITALISM, 244 ff.

WARD, J., 13 ff.

WHOLE, organic, 237 ff. ; types of, 162 ff.

WHOLE AND PART, organic, 422.

WHOLE-PART, 107.

WOODBRIDGE, F. J. E., 441.

WORDS, use of, 21 ff.

HPHE following pages contain advertisements of a
few of the Macmillan books on kindred subjects.

1

"Should be read by every thoughtful person in America"

The New Democracy

An Essay on Certain Political and Economic Tendencies
in the United States

BY WALTER E. WEYL, PH.D.

A chief issue in the Presidential Campaign is "Socialized Democ-
racy." Dr. Weyl's new work gives a clear summary of its causes

and aims. Crown 8vo, blue cloth, gilt top, $2.00 net; postpaid, $2.12

SOME PRESS COMMENTS

"A masterly, scathing, and absolutely fearless arraignment of
things that ought not to be in a republic, and of tendencies that no
democracy ought to tolerate." — Boston Herald.

" A thoughtful volume ... a big synthesis of the whole social
problem in this country. A keen survey." — Chicago Evening Post.

" A searching and suggestive study of American life. ... A book
to make people think. . . . Notable for its scholarship and brilliant
in execution, it is not merely for the theorist, but for the citizen."

— Newark Evening News.

"Dr. Weyl's book is a strong analysis of the whole subject. It will
be read widely and will exercise a large influence." — The Evening Mail.

" Dr. Weyl has read the modern writing on the wall and interprets
it to us more clearly than has yet been done." — New York Globe.

"A masterly interpretation of the industrial, political, social, and
moral revolution that is going on in this country." — Albany Argus.

" A complete and circumstantial statement of the whole case . . .
our social and economic unrest is not to lead to a war of classes, but
to a 'National readjustment.' " — New York Tribune.

" The best and most comprehensive survey of the general social and
political status and prospects that has been published of late years."

— The Pittsburgh Post.

THE MACMILLAN COMPANY

Publishers 64-66 Fifth Avenue New Tork

A LIST OF MACMILLAN PUBLICATIONS ON
PHILOSOPHY

A Student's History of Philosophy

By ARTHUR KENYON ROGERS, Professor of Philosophy in Butler College.
Published in New York, 1901. Second edition, 1907.

Cloth, 511 pages, 8vo, $2.00 net

A History of Philosophy

With Especial Reference to the Formation and Development of its Problems
and Conceptions. By Dr. W. WINDELBAND, Professor of Philosophy in the
University of Strassburg. Authorized translation by James H. Tufts, Ph.D.,
Professor of Philosophy in the University of Chicago. Published in New
York, 1893. Second edition, revised and enlarged, 1901. Latest reprint, 1907.

Cloth, 726 pages, 8vo, $4.00 net

An Introduction to Philosophy

By GEORGE STEWART FULLERTON Professor of Philosophy in Columbia Univer-
sity. Published in New York, 1906. Reprinted, 1908.

Cloth, 322 pages, I2mo, $1.60 net

The Persistent Problems of Philosophy

An Introduction to Metaphysics through the Study of Modern Systems. By
MARY WHITON CALKINS, Professor of Philosophy and Psychology in Wellesley
College. Published in New York, 1907. Second edition, 1908.

Cloth, 575 pages, 8vo, $2.50 net

Introduction to Philosophy

By WILLIAM JERUSALEM, Lecturer in Philosophy and Pedagogy at the Uni-
versity of Vienna. Translated from the fourth edition by Charles F. Sanders.
Published in New York, 1910. Cloth, 319 pages, izmo, $1.50 net

An Introduction to Systematic Philosophy

By WALTER T. MARVIN, Ph.D. Published in New York, 1003.

Cloth, 572 pages, 8vo, $3.00 net

A Brief Introduction to Modem Philosophy

By ARTHUR KENYON ROGERS, Ph.D. Published in New York, 1899. Latest
reprint, 1909. Cloth, 360 pages, i2mo, $1.25 net

English Philosophy

A Study of its Method and General Development. By THOMAS M. FORSYTE.
Published in London, 1910. Cloth, 231 pages, 8vo, $1.75 net

Dogmatism and Evolution

Studies in Modern Philosophy. By THEODORE DE LACUNA, Ph.D., Professor
of Philosophy in Bryn Mawr College, and GRACE ANDRUS DE LACUNA, Ph.D.
Published in New York, 1910. Cloth, 259 pages, 8vo, $1.75 net

The Evolution of Mind

By JOSEPH McCABE. Published in London, 1910.

Cloth, 287 pages, 8w, $2.00 net

Psychology: Normal and Morbid

By CHARLES A. MERCIER. Published in London, 1907.

Cloth, 518 Pages, 8vo, $4.00 net

Studies in Clinical Psychiatry

By LEWIS C. BRUCE, M.D., G.R.C.P.E. Published in London, 1906.

Cloth, 246 pages, 8vo, $4.00 net

Clinical Psychiatry

A Text-book for Students and Physicians. Abstracted and adapted from the
seventh German edition of Kraepelin's " Lehrbuch der Psychiatric." By A.
Ross DIEFENDORF. Published in New York, 1902. Second edition, 1907.

Cloth, 562 pages, 8vo, $3.75 net

Neurological and Mental Diagnosis

A Manual of Methods. By L. PIERCE CLARK, M.D. , Neurologist, Vanderbilt
Clinic, Columbia University, and A. Ross DIEFENDORF, M.D., Lecturer in
Psychiatry in Yale University. Published in New York, 1908.

Cloth, illustrated, 188 pages, i2mo, $1.25 net

The Major Symptoms of Hysteria

Fifteen Lectures Given in the Medical School of Harvard University. By
PIERRE JANET, M.D. Published in New York, 1907.

Cloth, 337 pages, izmo, $1.75 net

A Text-Book of Insanity

By CHARLES MERCIER. Published in London, 1902.

Cloth, 222 pages, I2tno, $1.75 net

The Animal Mind

A Text-book of Comparative Psychology. By MARGARET FLOY WASHBURN,
Ph.D., Associate Profe sor of Philosophy in Vassar College. Published in
New York, 1908. Cloth, 333 pages, I2tno, $1.60 net

The Dancing Mouse

A Study in Animal Behavior. By ROBERT M. YERKES, Ph.D., Instructor in
Comparative Psychology in Harvard University. Vol. I in the Animal Be-
havior Series. Published in New York, 1907.

Cloth, illustrated, 290 pages, I2mo, $1.25 net

./Esthetic Principles

By HENRY RUTGERS MARSHALL, M.A. Published in New York, 1895.
Latest reprint, 1901. Cloth, 201 pages, izmo, $1.25 net

Consciousness

By HENRY RUTGERS* MARSHALL. Published in New York and London, 1909.

Cloth, 686 pages, 8vo, $4.00 net
The Psychology of Religious Belief

By JAMES BISSETT PRATT, Ph.D. Published in New York, 1906. Reprinted
1907. Cloth, 327 pages, I2mo, $1.50 net

The World and the Individual

Gifford Lectures delivered at University of Aberdeen. By JOSIAH ROYCE,
Ph.D., Professor of the History of Philosophy in Harvard University.
Part I, The Four Historical Conceptions of Being.

Cloth, 588 pages, I2mo, $3.00 net
Part II, Nature, Man, and the Moral Order.

Cloth, 480 pages, I2m>p, $2.25 net
Published in New York, 1899. Third reprint, 1908.

The World a Spiritual System

An Outline of Metaphysics. By JAMES H. SNOWDEN. Published in New
York, 1910. Cloth, 316 pages, I2mo, $1.50 net

Modern Thought and the Crisis in Belief

By R. M. WENLEY, of the University of Michigan. Published in New York,
1910. Cloth, 364 pages, izmo, $1.50 net

From Epicurus to Christ

A Study in the Principles of Personality. By WILLIAM DE WITT HYDE, Presi-
dent of Bowdoin College. Published in New York, 1904. Third reprint, 1908.

Cloth, 285 pages, I2mo, $1.50 net

The Religious Conception of the World

An Essay in Constructive Philosophy. By ARTHUR KENYON ROGERS, Ph.D.,
Professor of Philosophy in Butler College. Published in New York, 1907.

Cloth, 284 pages, i2mo, $1.50 net

The Library of Philosophy

Edited by J. H. MUIRHEAD, LL.D.

Erdmann. History of Philosophy. Three Volumes. Cloth, 8vo, $10.50 net
Bosanqnet. History of ^Esthetic. Cloth, 8w, $2.75 net

Pfleiderer. Development of Rational Theology in Germany and in Great

Britain since Kant. Cloth, 8vo, $2.75 net

Bonar. Philosophy and Political Economy in Some of their Historical

Relations. Cloth, 8vo, $2.75 net

Bradley. Appearance and Reality. Cloth, 8w, $2.75 net

Ritchie. Natural Rights. Cloth, 8w, $2.75 net

Sigwart. Logic : Two volumes — Vol. I, The Judgment, Concept, and

Inference. Vol. II, Logical Methods. Cloth, 8w, $5.50 net

Stout. Analytic Psychology. Third edition. Two volumes.

Cloth, 8vo, $5.50 net

Albee. History of English Utilitarianism. Cloth, 8w, $2.75 net

Villa. Contemporary Psychology. Cloth, 8vo, $2.75 net

Baldwin. Thought and Things. Cloth, 8vo, two vols., each $2.73 net

Urban. Valuation : its Nature and Laws. Cloth, 8vo, $2.75 net

Pillsbury. Attention. Cloth, 8vo, $2.75 net

Bergson. Time and Free Will. Matter and Memory. Cloth, 8w, each $2.75 net
Baillie. HEGEL'S Phenomenology of Mind. Cloth, 8vo, twovols., each $2.75 net

PSYCHOLOGY

A Text-book of Psychology

By EDWARD BRADFORD TITCHENER. Part I. Published in New York, 1896,
as "An Outline of Psychology." Published under present title, 1909.

Cloth, 311 pages, izmo, $1.30 net

An Outline of Psychology

By EDWARD BRADFORD TITCHENER. Published in New York, 1896. Ten re-
prints. New edition with additions, 1908. Cloth, 379 Pages, izmo, $1.50 net

A Primer of Psychology

By EDWARD BRADFORD TITCHENER. Published in New York, 1898. Second
edition revised, 1899. Seventh reprint, 1906. Cloth, 316 pages, izmo, $1.00 net

A First Book in Psychology

By MARY WHITON CALKINS, Professor of Philosophy and Psychology in
Wellesley College. Published in New York, 1910.

Cloth, 424 pages, izmo, $1.90 net

An Introduction to Psychology

By MARY WHITON CALKINS, Professor of Philosophy and Psychology in
Wellesley College. Published in New York, 1901. Latest reprint, 1905.

Cloth, 512 pages, I2mo, $2.00 net

Outlines of Psychology

An Elementary Treatise, with Some Practical Applications. By JOSIAH
ROYCE, Ph.D., LL.D., Professor of the History of Philosophy in Harvard
University. Published in New York, 1903. Latest reprint, 1906.

Cloth, 392 pages, izmo, $1.00 net

Experimental Psychology

A Manual of Laboratory Practice. By EDWARD BRADFORD TITCHENER.
Vol. I, Qualitative Experiments. Part I, Student's Manual.

Cloth, 214 pages, 8vo, $1.60 net

Part II, Instructor's Manual. Cloth, 456 pages, 8w, $2.50 net

Published in New York, 1901.
Vol. II, Quantitative Experiments. Part I, Student's Manual.

Cloth, 208 pages, 8w, $1.40 net

Part II, Instructor's Manual. Cloth, 453 pages, 8vo, $2.50 net

Published in New York, 1905.

Experimental Psychology and Its Bearing upon Culture

By GEORGE MALCOLM STRATTON, Associate Professor of Psychology in the
University of California. Published in New York, 1903.

Cloth, 331 pages, izmo, $2.00 net

Genetic Psychology

By EDWIN A. KIRKPATRICK. Published in New York, 1909.
1886-7 Cloth, 373 pages, izmo, $1.25 net

The Principles of Religious Development

A Psychological and Philosophical Study. By GEORGE GALLOWAY. Pub-
lished in London, 1909. Cloth, 362 pages, 8w, $3.00 net

Social Psychology

An Outline and Source Book. By EDWARD ALSWORTH Ross, Professor of
Sociology in the University of Wisconsin. Published in New York, 1908.

Cloth, 372 pages, izmo, $1.50 net

Mind in Evolution

By L. T. HOBHOUSE. Published in London, 1901.

Cloth, 406 pages, 8vo, $3.25 net

Hume : the Relation of the Treatise of Human Nature

Book I to the Inquiry Concerning Human Understanding. By W. B. ELKIN,
Ph.D. Published in New York, 1904. Cloth, 330 pages, i2mo, $1.50 net

The Psychology of Thinking

By IRVING ELGAR MILLER, Ph.D. Published in New York, 1909.

Cloth, 303 pages, izmo, $1.25 net

Experimental Psychology of the Thought Processes

By EDWARD BRADFORD TITCHENER, Sage Professor of Psychology at Cornell
University. Published in New York, 1909. Cloth, 308 pages, $1.25 net

Lectures on the Elementary Psychology of Feeling and Attention

By EDWARD BRADFORD TITCHENER. Published in New York, 1908.

Cloth, 404 pages, izmo, $1.40 net

Instinct and Reason

An Essay concerning the Relation of Instinct to Reason, with Some Special
Study of the Nature of Religion. By HENRY RUTGERS MARSHALL, M.A.
Published in New York, 1898. Cloth, 573 pages, 8w, $3.50 net

Structure and Growth of the Mind

By W. MITCHELL. Published in London, 1907. Cloth, 512 pages, 8vo, $2.60 net

Why the Mind has a Body

By C. A. STRONG, Professor of Psychology in Columbia University. Pub-
lished in New York, 1903. Reprinted, 1908. Cloth, 355 pages, 8vo, $2.50 net

Outlines of Metaphysics

By JOHN S. MACKENZIE, M.A., Glasgow ; Litt.D., Cambridge. Published in
London, 1902. Second edition, 1006. Cloth, 175 pages, i2mo, $1.10 net

A System of Metaphysics

By GEORGE STEWART FULLERTON, Professor of Philosophy in Columbia Uni-
versity. Published in New York, 1004. Cloth, 627 pages, 8vo, $4.00 net

Elements of Metaphysics

By A. E. TAYLOR, Professor of Philosophy in McGill University, Montreal.
Published in New York, 1007. Cloth, 419 pages, 8vo, $2.60 net

Concepts of Philosophy

In Three Parts. Part I, Analysis. Part II, Synthesis, a. From Physics to
Sociality, b. From Sociality to Religion. Part III, Deductions. By ALEX-
ANDER THOMAS ORMOND. Published in New York, 1906.

Cloth, 722 pages, 8vo, $4.00 net
The Problems of Philosophy

By HARALD HOFFDING. Translated by Galen M. Fisher. With a preface by
William James. Published in New York, 1005. Reprinted, 1906.

Cloth, 201 pages, idmo, $1.00 net
What is Pragmatism ?

By JAMES BISSETT PRATT, Ph.D., Assistant Professor of Philosophy in Wil-
liams College. Published in New YorkjiQOQ. Cloth, 256 pages, izmo, $1.25 net

Studies in Humanism

By F. C. S. SCHILLER, M.A., D.Sc. Published in London, 1907.

Cloth, 492 pages, 8vo, $3.25 net

Riddles of the Sphinx

A Study in the Philosophy of Humanism. By F. C. S. SCHILLER. Published
in London, 1910. Cloth, 478 pages, 8vo, $3.00 net

The Critical Philosophy of Immanuel Kant

By EDWARD CAIRO, LL.D., Professor of Moral Philosophy in the University
of Glasgow. Published in Glasgow, 1889. Second edition 1008.

Two volumes. Cloth, 8vo
Vol. I, 654 pages. Vol. II, 660 pages. The set, $6.25 net

The Philosophy of Kant Explained

By JOHN WATSON, M.A., LL.D., Professor of Moral Philosophy in the Uni-
versity of Queen's College, Kingston, Canada. Published in Glasgow, 1908.

Cloth, 515 pages, 8vo, $3.75 net

The Philosophy of Kant

As contained in Extracts from His Own Writings. Selected and translated by
JOHN WATSON, LL.D., Professor of Moral Philosophy in the University of
Queen's College, Kingston, Canada. Published in Glasgow, 1888. Second
edition, 1908. Cloth, 356 pages, I2mo, $1.75 net

THE MACMILLAN COMPANY

Publishers 64-66 Fifth Avenue New York

THE LIBRARY
UNIVERSITY OF CALIFORNIA

Santa Barbara

THIS BOOK IS DUE ON THE LAST DATE
STAMPED BELOW.

0000599563